Brief Bio

Mahdi Aliyari Shoorehdeli received his B.S. degree in Electronic Engineering in 2001, his M.Sc and Ph.D. degrees in Control Engineering in 2003 and 2008 in Control Engineering from K. N. Toosi University of Technology, Tehran repetitively. He has been with the Mechatronics Department of K. N. Toosi University of Technology, Tehran, Iran since 2010. His research interests include Fault Diagnosis, System Identification and Alarm Management.

  • Doctor of Philosophy, Electrical Engineering-Control Systems, 2003- 2007

    K. N. Toosi University of Technology, Tehran, Iran

    Thesis:” Stability Analysis of Neuro-Fuzzy Systems based on Hybrid Training Algorithms,

    Supervisor: Professor Mohammad Teshnelab

    Advisor: Professor Ali Khaki Sedigh

  • Master of Science, Electrical Engineering-Control Systems, 2001-2003

    K. N. Toosi University of  Technology, Tehran, Iran

    Thesis:” Prediction of Air Pollution Dynamics in Arak by Using Neural Networks”,

    Supervisor: Professor Ali Khaki Sedigh

  • Bachelor of Science, Electrical Engineering -Electronics, 1997-2001

    K. N. Toosi University of  Technology, Tehran, Iran

    Thesis:” Digital Oscilloscope“,

    Supervisor: Dr. Abedi

Research

  • Fault Diagnosis and Fault Tolerant Control
  • Industrial System Identification
  • Multi objective Optimizations and Convergence Analysis
  • Neuro- fuzzy systems stability and control
  • Fuzzy systems and control
  • Nonlinear Control

Experience

Intelligent control and multi objective optimization (2006-6 Months)

Control Eng. (Emmy Johnson) Building, University of Sheffield, Sheffield, England.

Supported By: British Council

Supervisor: Professor M. Mahfouf

Hybrid Controls Based on Intelligent and Classic Controllers (2005-2008)

Intelligent Systems Laboratory, ISLAB Electrical eng., K. N Toosi University of Technology,

Supported By: K. N Toosi University of Technology, Grant Projects.

Identification of Tokamak Systems (2009-2011)

Advanced Process Automation and Control (APAC) Research Group,

Supported By: K. N Toosi University of Technology.

Design and Implementation of Electromechanical Actuators (2012-2015)

K. N. Toosi university of Technology and Maham Company.

Supported By: Maham Company.

Design and Implementation of Electro hydrostatic Actuators (2012-2015)

K. N. Toosi university of Technology and Maham Company.

Supported By: Maham Company.

Gas Turbine Fault Detection based on Signal based Methods (2014-now)

K. N. Toosi university of Technology and MAPNA – MCCO.

Supported By: MAPNA Company.

Publication

Abstract:TitleYearType
Abstract:

This paper deals with the design of adaptive fuzzy dynamic surface control for uncertain strict-feedback nonlinear systems with asymmetric time-varying output constraints in the presence of input saturation. To approximate the unknown nonlinear functions and overcome the problem of explosion of complexity, a Fuzzy logic system is combined with the dynamic surface control in the backstepping design technique. To ensure the output constraints satisfaction, an asymmetric time-varying Barrier Lyapunov Function (BLF) is used. Moreover, by applying the minimal learning parameter technique, the number of the online parameters update for each subsystem is reduced to 2. Hence, the semi-globally uniformly ultimately boundedness (SGUUB) of all the closed-loop signals with appropriate tracking error convergence is guaranteed. The effectiveness of the proposed control is demonstrated by two simulation examples.

Adaptive fuzzy dynamic surface control of nonlinear systems with input saturation and time-varying output constraints
L Edalati, A Khaki Sedigh, M Aliyari Shooredeli, A Moarefianpour
Mechanical Systems and Signal Processing
2018Journal
Abstract:

This paper investigates the condition of polyethylene (PE) pipelines as a case study. This study introduces a novel method to detect and diagnose defects of high-density polyethylene (HDPE) pipes. The pipe defect detector technique (PDDT) is designed to capture and process the images from the inner surface of pipes. Consequently, PDDT is one of the nondestructive ways to investigate possible defects in pipes. The PDDT’s outcome offers valuable information regarding the shape, orientation, and length of defects in the inner surface of the pipe. This information plays an important role in defining the lifetime of the pipe and fault prediction. In this paper, a database consisting of a total 350 images was used to train, test, and verify a neural network system. For this purpose, input image quality was enhanced by applying Gabor and entropy filters. Then, the trained neural network was used to classify the input images into five defect categories. These categories are defined in a way to describe the shape and the orientation of the defects. Afterward, a curve completion method (CMM) that effectively derives the defect dimensions such as diameter and length was introduced. Finally, the life prediction methods that can use PDDT’s result to predict the time that actual fault may occur in the pipe are discussed.

Detection and Isolation of Interior Defects Based on Image Processing and Neural Networks: HDPE Pipeline Case Study
Shiva Safari, Mahdi Aliyari Shoorehdeli
Journal of Pipeline Systems Engineering and Practice
2018Journal
Abstract:

Fault detection in non‐linear system has drawn a lot of attention recently. A typical solution is the generalization of linear methods to include non‐linear dynamics. This study addresses fault detection in non‐linear systems by extending parity relations using Takagi‐Sugeno (T.S) fuzzy models. Parity equations for linear systems are a residual generation method that has appealing capabilities in fault detection. T.S fuzzy systems are also extensively used in modelling of non‐linear systems. In this paper, parity equations are rewritten in the form of non‐linear systems that can be modelled by T.S fuzzy system. An advantage of this approach is that parity vector can be derived from relations explicitly. An algorithm is proposed to show how a residual can be generated in this manner. Simulation results on the fault detection of a mass‐spring‐damper system show the effectiveness of the proposed method.

Generalization of parity space to fault detection based on Takagi‐Sugeno fuzzy models for non‐linear dynamic systems
Majid Ghaniee Zarch, Mahdi Aliyari Shoorehdeli
Expert Systems
2018Journal
Abstract:

Decision-making systems are known as the main pillar of industrial alarm systems, and they can directly effect on system's performance. It is evident that because of hidden attributes in the measurements such as correlation and nonlinearity, thresholding systems faced wrong separation defining by Missed Alarm Rate (MAR) and False Alarm Rate (FAR). This study introduced a novel extended adaptive thresholding based on mean-change point detection algorithm and shows that it is more efficient than other existing thresholding algorithm in the literature. Number hypothetical and industrial examples are given to delineate the capabilities and limitation of proposed method and prove its effectiveness in an industrial alarm system.

A novel extended adaptive thresholding for industrial alarm systems
Mahdi Bahar-Gogani, Koorosh Aslansefat, Mahdi Aliyari Shoorehdeli
Electrical Engineering (ICEE), 2017 Iranian Conference on
2017Conference
Abstract:

In this paper, a recurrent neural network coupled with Kalman filter is proposed to identify dynamic terms of robotic manipulator. By cooperating some inherent characteristics of robot, this network has the capability to individually identify nonlinear terms using Weighted Augmentation Error (WAE). To present the infrastructure of architecture, an adaptive scheme based on the conventional Back Propagation (BP) is firstly driven using the Gradient Descent (GD) method. Additionally, a stable adaptive updating rule is extracted from the discrete time Lyapunov candidate as an approach for the general nonlinear system identification. Then, this approach is applied to the predefined network. To experimentally validate the computational efficiency and control applicability of the proposed method, Adaptive Neural Network Based Inverse Dynamic Control (ANN-Based-IDC) is employed on a laboratory-scaled twin-rotor CE-150 helicopter. This experiment illustrates enhancement of steady-state performance from 2-to-3 times more in compared with simple PID. Moreover, disturbance rejection and robustness tests admit capability of the method for online dynamic identification in the presence of output and dynamic perturbation.

Adaptive recurrent neural network with Lyapunov stability learning rules for robot dynamic terms identification
Pedram Agand, Mahdi Aliyari Shoorehdeli, Ali Khaki-Sedigh
Engineering Applications of Artificial Intelligence
2017Journal
Abstract:

In the recent years, artificial neural network have been used to improvement of system identification. The performance of neural network directly depends on the hidden layer, which include weights and activation functions of the network. In addition Genetic Algorithms are used to learn of neural network as a type of evolutionary computing algorithms. In this paper, the structure of hidden layers and weights are modified by using biological neuron model of Izhikevich. These two methods, Genetic Algorithms and biological model of neuron, merge together for designing a novel structure.

An artificial neural network based on Izhikevich neuron model
Katayoon Taherkhani, Mahdi Aliyari Shoorehdeli
Electrical Engineering (ICEE), 2017 Iranian Conference on
2017Conference
Abstract:

In this letter, we point out that the asymptotic convergence, claimed in Theorem 2, of the output residual and parameter estimation error after fault occurrence are guaranteed by the performance of the fault diagnosis observer is not quite right. The proof of the asymptotic convergence is contributed by Lemma 1 and negative semidefiniteness of the first difference of Lyapunov candidate function. Here, it is shown that utilizing Lemma 1 yields in some disputed points in the proof of Theorem 2. On the other hand, the proof of Theorem 2 is not mathematically correct. Therefore, the guarantee of the asymptotic convergence mentioned for FD observer in Theorem 2 is not realizable.

Comments on “A Novel Fault Diagnostics and Prediction Scheme Using a Nonlinear Observer With Artificial Immune System as an Online Approximator”
L Mahmoodi, M Aliyari Shoorehdeli
IEEE Transactions on Control Systems Technology
2017Journal
Abstract:

In This study, we present a new sensor fault detection approach based on nonlinear parity technique in presence of sensor noise. Conventionally analytical redundancy (AR) was used to fault detection and isolation in linear systems. The proposed parity space approach with nonlinear analytical redundancy (NLAR) technique can be applied to detect sensor fault in the nonlinear affine systems with mentioned class. The proposed approach will be implemented in pH neutralization system. At the end nonlinear fault detection and identification algorithm will be successfully implemented, examined and reported.

Nonlinear parity approach to sensor fault detection in pH neutralization system
Hamed Tolouei, Mahdi Aliyari Shoorehdeli
Electrical Engineering (ICEE), 2017 Iranian Conference on
2017Conference
Abstract:

A spatially-constrained clustering algorithm is presented in this paper. This algorithm is a distributed clustering approach to fine-tune the optimal distances between agents of the system to strengthen the data passing among them using a set of spatial constraints. In fact, this method will increase interconnectivity among agents and clusters, leading to improvement of the overall communicative functionality of the multi-robot system. This strategy will lead to the establishment of loosely-coupled connections among the clusters. These implicit interconnections will mobilize the clusters to receive and transmit information within the multi-agent system. In other words, this algorithm classifies each agent into the clusters with the lowest cost of local communication with its peers. This research demonstrates that the presented decentralized method will actually boost the communicative agility of the swarm by probabilistic proof of the acquired optimality. Hence, the common assumption regarding the full-knowledge of the agents’ primary locations has been fully relaxed compared to former methods. Consequently, the algorithm’s reliability and efficiency is confirmed. Furthermore, the method’s efficacy in passing information will improve the functionality of higher-level swarm operations, such as task assignment and swarm flocking. Analytical investigations and simulated accomplishments, corresponding to highly-populated swarms, prove the claimed efficiency and coherence.

Optimal distributed interconnectivity of multi-robot systems by spatially-constrained clustering
Mahdi Aliyari Sh Matin Macktoobian
Adaptive Behavior
2017Journal
Abstract:

In this study, a novel robust fault diagnosis scheme is developed for a class of nonlinear systems when both fault and disturbance are considered. The proposed scheme includes both component and sensor fault with nonlinear system that transferred to nonlinear Takagi-Sugeno (T-S) model. It considers a larger category of nonlinear system when fuzzification is used for only nonlinear distribution matrices. In fact the proposed method covers nonlinear systems could not transform to linear T-S model. This paper studies the problem of robust fault diagnosis based on two fuzzy nonlinear observers, the first one is a fuzzy nonlinear unknown input observer (FNUIO) and the other is a fuzzy nonlinear Luenberger observer (FNLO). This approach decouples the faulty subsystem from the rest of the system through a series of transformations. Then, the objective is to design FNUIO to guarantee the asymptotic stability of the error dynamic using the Lyapunov method; meanwhile, FNLO is designed for faulty subsystem to generate fuzzy residual signal based on a quadratic Lyapunov function and some matrices inequality convexification techniques. FNUIO affects only the fault free subsystem and completely removes any unknown inputs such as disturbances when residual signal is generated by FNLO is affected by component or sensor fault. This novelty and using nonlinear system in T-S model make the proposed method extremely effective from last decade literature. Sufficient conditions are established in order to guarantee the convergence of the state estimation error. Thus, a residual generator is determined on the basis of LMI conditions such that the estimation error is completely sensitive to fault vector and insensitive to the unknown inputs. Finally, an numerical example is given to show the highly effectiveness of the proposed fault diagnosis scheme.

Robust fault diagnosis scheme in a class of nonlinear system based on UIO and fuzzy residual
S Hamideh Sedigh Ziyabari, Mahdi Aliyari Shoorehdeli
International Journal of Control, Automation and Systems
2017Journal
Abstract:

In this study, a novel fuzzy unknown input observer for robust fault estimation scheme is developed when both faults and unknown input are considered. The proposed scheme includes component fault with nonlinear distribution matrix in state equation, unknown input signal in state and output equations. After that, Takagi-Sugeno (T-S) model is used to create multiple models. While T-S model is used for only the nonlinear distribution matrix of the fault signal, a larger category of nonlinear system will be included. Two set of observers are considered, the first one is extended fuzzy unknown input observer (EFUIO) and the other one is fuzzy sliding mode observer (FSMO). The approach decoupled the faulty subsystem from the rest of the system through a series of linear transformations. Then, the objective is to design EFUIO to guarantee the asymptotic stability of the error dynamic using the Lyapunov method. Unknown input is removed; meanwhile, FSMO is designed for faulty subsystem to guarantee estimation of fault. Sufficient conditions are established in order to guarantee the convergence of the state estimation error and the results are formulated in the form of linear matrix inequalities (LMIs). Finally, a simulation study on an electromagnetic suspension system (EMS) is presented to demonstrate the performance of the results compared with a pure SMO.

Robust fuzzy fault estimation based on decoupled transform and unknown input sliding mode observer
S Hamideh Sedigh Ziyabari, Mahdi Aliyari Shoorehdeli
Electrical Engineering (ICEE), 2017 Iranian Conference on
2017Conference
Abstract:

In this paper, model based fault detection of gas turbine using linear and non-linear methods (multilayer perceptron and radial basis function neural network models) is studied. We contemplate IGV positions and gas flow as input and sensors related to compressor as outputs. Then residual signals will be obtained based on system model. In addition, by these signals and exert the fixed and adaptive thresholds, the fault occurred in the V94. 2 gas turbine which is pollution of vane compressor (Fouling detection) has identified and diagnosed. Consequently, by comparing the obtained results from different fault detection methods, we determine the most appropriate signal output that led to better and reliable result. All simulations have been carried out by using real data taken from an V94. 2 industrial gas turbine 927 power plant in Fars.

V94. 2 industrial gas turbine compressor fouling detection based on system identification methods, neural networks and experimental data
Sahar Rahimi Malekshan, Mahdi Aliyari Shoorehdeli, Mostafa Yari
Electrical Engineering (ICEE), 2017 Iranian Conference on
2017Conference
Abstract:

Filtering is an effective method of alarm management family that can reduce false and missed alarm rates significantly. Simple and effective techniques of fault diagnosis methods are popular in industry. So, deriving a simple analytic filter design approach is important. This study proposes a simple analytic linear filter design based on a probabilistic model of the system. At last, the effectiveness of the proposed method is showed in the deposition fault detection of a V94. 2 gas turbine with 162.1 MW and 50 Hz as the nominal power and frequency respectively. It is built by MAPNA group (originally built by SIEMENS) and set up in Shiraz power plant, Shiraz city of Iran.

Alarm management based fault diagnosis of V94. 2 gas turbines by applying linear filters
Hamid Alikhani, Mahdi Aliyari Shoorehdeli, Mostafa Yari
Robotics and Mechatronics (ICROM), 2016 4th International Conference on
2016Conference
Abstract:

In this study, a novel fuzzy robust fault estimation scheme is developed for a class of nonlinear systems when both fault and disturbance are considered. The proposed scheme includes component fault with a nonlinear distribution matrix; as a result, the Takagi–Sugeno model is used to create multiple models. While the Takagi–Sugeno model is used for only the nonlinear distribution matrix of the fault signal, a larger category of nonlinear systems will be considered. This paper presents the problem of robust fault estimation based on fuzzy nonlinear observers, the first one is a fuzzy unknown input observer and the other one is a fuzzy sliding mode observer. The approach decoupled the faulty subsystem from the rest of the system through a series of transformations. Then, the objective is to design a fuzzy unknown input observer guaranteeing the asymptotic stability of the error dynamic using the Lyapunov method and completely removing disturbances; meanwhile, a fuzzy sliding mode observer is designed for a faulty subsystem to generate an estimation of fault based on a quadratic Lyapunov function and some matrices inequality convexification techniques. The sliding motion affects only the faulty subsystem through a novel reduced order fuzzy sliding mode observer; meanwhile, all disturbances are completely removed by fuzzy unknown input observer. Sufficient conditions are established in order to guarantee the convergence of the state estimation error and the results are formulated in the form of linear matrix inequalities. Thus, an exact fault estimator is determined on the basis of linear matrix inequality conditions while the estimation fault is completely insensitive to the disturbance. Finally, a simulation study on an electromagnetic suspension system is presented to demonstrate the g performance of the results compared with a pure sliding mode observer.

Fuzzy robust fault estimation scheme for a class of nonlinear systems based on an unknown input sliding mode observer
Hamideh Sedigh Ziyabari, Mahdi Aliyari Shoorehdeli
Journal of Vibration and Control
2016Journal
Abstract:

This paper presents a Gaussian radial basis function neural network based on sliding mode control for trajectory tracking and vibration control of a flexible joint manipulator. To study the effectiveness of the controllers, designed controller is developed for tip angular position control of a flexible joint manipulator. The adaptation laws of designed controller are obtained based on sliding mode control methodology without calculating the Jacobian of the flexible joint system. Also in this study, the anti-control is applied to reduce the deflection angle of flexible joint system. To achieve this goal, the chaos dynamic must be created in the flexible joint system. So, the flexible joint system has been synchronized to chaotic gyroscope system. In this study, control and anti-control concepts are applied to achieve the high quality performance of flexible joint system. It is tried to design a controller which is capable to satisfy the control and anti- control aims. The performances of the proposed control are examined in terms of input tracking capability, level of vibration reduction and time response specifications. Finally, the efficacy of the proposed method is validated through experimentation on QUANSER’s flexible-joint manipulator.

Hybrid Concepts of the Control and Anti-Control of Flexible Joint Manipulator
Mojtaba Rostami Kandroodid, Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Maysam Zamani Pedram
International Journal of Robotics, Theory and Applications
2016Journal
Abstract:

This study presents a novel indirect adaptive hierarchical fuzzy sliding mode controller for a class of high-order SISO nonlinear systems in normal form with unknown functions in the presence of bounded disturbance. The hierarchical fuzzy system is able to reduce the number of rules and parameters with respect to ordinary fuzzy systems. On-line tuning algorithm for consequent part parameters of fuzzy rules in different layer of hierarchical fuzzy system is derived using defined Lyapunov function. Two theorems are proved to show that the suggested adaptive schemes can achieve asymptotically stable tracking of a reference input with guarantee of the bounded system signals. One for unity control gain and the other for non-unity control gain. To show the effectiveness of the proposed method, control of three systems are considered in the simulations. The simulations results demonstrate the validity, efficiency and feasibility of the proposed approach to control of nonlinear dynamic system.

Indirect adaptive hierarchical fuzzy sliding mode controller for a class of nonlinear systems
MA Shoorehdeli M Mansouri, M Teshnehlab
Journal of Intelligent & Fuzzy Systems
2016Journal
Abstract:

In this contribution, full probability distribution of parameters of ARX model is obtained for on-line problems by means of Bayesian approach and Markov chain Monte Carlo method (MCMC), which provides the ability to be applied on time-varying ARX models as well. Full probability distribution of parameters represent whole available knowledge of parameters. So, decision makers can follow any policies to make decision about point estimation, like dynamic point estimation. Moreover, the Bayesian approach has great potential in combining sources of knowledge much more easier. To decrease the computational efforts, full probability of model parameters are updated based on size-varying partitions. Furthermore, incorporating the posterior probability of previous partition into the jump probability of current partition, in MCMC method, improves the performance of the proposed algorithm from the computation and convergence rate point of view. Simulation results demonstrate the effectiveness and validity of the proposed algorithm.

On-line Full Probability Distribution Identification of ARX Model Parameters Based on Bayesian Approach
Amir HoseinValadkhani, Aminollah Khormali, Mahdi Aliyari Shoorehdeli Hamid Khaloozadeh Alireza Fatehi
IFAC-PapersOnLine
2016Journal
Abstract:

In this paper, particle dynamics and stability analysis of gravitational search algorithm (GSA) are investigated. The GSA is a swarm optimization algorithm which is inspired by the Newtonian laws of gravity and motion. Previously, the convergence analysis of the GSA and improved GSA algorithms were presented to demonstrate each particle converges. In this study, the stability of the particle dynamics using Lyapunov stability theorem and the system dynamics concept is analyzed. Sufficient conditions of stability analysis are investigated and utilized for adapting parameters of the GSA. The modified algorithm based on stability analysis is compared with the standard GSA, PSO, RGA, and two methods of improved GSA in terms of average, median, and standard deviation of best-so-far solutions. Simulation results demonstrate the validity and feasibility of the proposed modified GSA. In solving the optimization problem of various nonlinear functions, the high performance is achieved.

Stability analysis of particle dynamics in gravitational search optimization algorithm
Faezeh Farivar, Mahdi Aliyari Shoorehdeli
Information Sciences
2016Journal
Abstract:

In this paper, a novel scheme is presented to conquer the motion-planning problem for autonomous space robots. Minimizing the consumed energy of atomic batteries within the daily planetary missions of robot on the planet is taken into account, i.e., utilization of the generated solar power by its embedded photocells leads to saving energy of batteries for night missions. Aforementioned objective could be acquired by appropriate interaction of motion planning paradigm with shadows of obstacles. Modeling of the shadow with the proposed artificial potential field leads to generalize the concept of potential fields not only for static and dynamic obstacles but also for being confronted with the intrinsic time-variant phenomena such as shadows. With due attention to the noticeable computational complexity of the introduced strategy, fuzzy techniques are applied to optimize the sampling times effectively. To accomplish this objective, a smart control scheme based on the fuzzy logic is mounted to the primitive version of algorithm. Regarding the need to identify some structural parameters of obstacles, PIONEER™ mobile robot is designed as a test bed for the verification of simulated results. Investigation on empirical accomplishments shows that the goal-oriented definition of Time–Variant Artificial Potential Fields is able to resolve the motion-planning problem in planetary applications.

Time-variant artificial potential field (TAPF): a breakthrough in power-optimized motion planning of autonomous space mobile robots
Matin Macktoobian, Mahdi Aliyari Shoorehdeli
Robotica
2016Journal
Abstract:

In this paper, a novel architecture in multilayer perceptron (MLP) neural network with flexible activation function and adaptive learning rate is presented for a data-driven identification of robot dynamics. It is assumed that the measurement of robot end-effector position, velocity and acceleration are available corrupted by Gaussian noise. Since some general property of robot dynamics are included in the proposed structure as well as optimization indices, this structure is envisaged having good performance in confronting with uncertainty in measurements. The main contribution of this paper is to propose a transparent neural network structure for identification of dynamic terms by introducing a gray-box identifier. Simulation results on 2-DOF serial manipulator reveal the accuracy of the method. Finally, experimental results on a laboratory-scaled twin rotor CE 150 helicopter indicate the applicability of the proposed method.

Transparent and flexible neural network structure for robot dynamics identification
Pedram Agand, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Electrical Engineering (ICEE), 2016 24th Iranian Conference on
2016Conference
Abstract:

In this paper, a novel adaptive hierarchical fuzzy control system based on the variable structure control is developed for a class of SISO canonical nonlinear systems in the presence of bounded disturbances. It is assumed that nonlinear functions of the systems be completely unknown. Switching surfaces are incorporated into the hierarchical fuzzy control scheme to ensure the system stability. A fuzzy soft switching system decides the operation area of the hierarchical fuzzy control and variable structure control systems. All the nonlinearly appeared parameters of conclusion parts of fuzzy blocks located in different layers of the hierarchical fuzzy control system are adjusted through adaptation laws deduced from the defined Lyapunov function. The proposed hierarchical fuzzy control system reduces the number of rules and consequently the number of tunable parameters with respect to the ordinary fuzzy control system. Global boundedness of the overall adaptive system and the desired precision are achieved using the proposed adaptive control system. In this study, an adaptive hierarchical fuzzy system is used for two objectives; it can be as a function approximator or a control system based on an intelligent-classic approach. Three theorems are proven to investigate the stability of the nonlinear dynamic systems. The important point about the proposed theorems is that they can be applied not only to hierarchical fuzzy controllers with different structures of hierarchical fuzzy controller, but also to ordinary fuzzy controllers. Therefore, the proposed algorithm is more general. To show the effectiveness of the proposed method four systems (two mechanical, one mathematical and one chaotic) are considered in simulations. Simulation results demonstrate the validity, efficiency and feasibility of the proposed approach to control of nonlinear dynamic systems.

Adaptive variable structure hierarchical fuzzy control for a class of high-order nonlinear dynamic systems
Mahdi Aliyari Shoorehdeli Mohammad Mansouri, Mohammad Teshnehlab
ISA transactions
2015Journal
Abstract:

In this study, a new strategy for fault detection and isolation is presented. This strategy is based on the design of a Lüneburg observer which is implemented via pole placement using linear matrix inequalities. Two residuals are formulated based on the state estimation error in order to be utilized in detecting and isolating faults happened on the system. Fault detection problem solves by changes occur in the residual value and fault isolation is done through determining threshold on residuals according to system behavior in faulty condition. The procedure performs in four simulations steps in which there are certain numbers of faults happen in the system in each step. This method is validated in simulation on a quadruple tank process while each faulty condition is considered as a leak at the bottom of a tank in the process. This can lead to an undesirable flow of liquid out of the tank which results to a decrease in tank's level. The simulation results represented in the paper shows the applicability of this strategy.

An observer based fault detection and isolation in quadruple-tank process
Zahra Gharaee, Mahdi Aliyari Shoorehdeli
Control and Decision Conference (CCDC), 2015 27th Chinese
2015Conference
Abstract:

The growing availability of high-resolution satellite imagery provides an opportunity for identifying road objects. Most studies associated with road detection are scene-related and also based on the digital number of each pixel. Because images can provide more details (including color, size, shape, and texture), object-based processing is more advantageous. Therefore, in this paper, to handle the existing uncertainty of satellite image pixel values, using type-2 fuzzy set theory in combination with object-based image analysis is proposed. Because the main challenges of the type-2 fuzzy set are parameter tuning and extensive computations, a hybrid genetic algorithm (GA) consisting of Pittsburgh and cooperative-competitive learning schemes is proposed to address these problems. The most prominent feature of our research in this work is to establish a comprehensive object-based type-2 fuzzy logic system that enables us to detect roads in high-resolution satellite images with no training data. The validation assessment of road detection results using the proposed framework for independent images demonstrates the capability and efficiency of our method in identifying road objects. For more evaluation, a type-1 fuzzy logic system with the same structure as type-2 is tuned. Evaluations show that type-1 fuzzy logic system quality in training is very similar to that of the proposed type-2 fuzzy framework. However, in general, its lower accuracy, as inferred by validation assessments, makes the type-1 fuzzy logic system significantly different from the proposed type-2.

Designing a new framework using type-2 FLS and cooperative-competitive genetic algorithms for road detection from IKONOS satellite imagery
Maryam Nikfar, Mohammad Javad Valadan Zoej, Mehdi Mokhtarzade, Mahdi Aliyari Shoorehdeli
Remote Sensing
2015Journal
Abstract:

This study presents fault detection of a heavy duty V94. 2 gas turbine which has 162.1 MW nominal power and 50 Hz nominal frequency and is located at Pareh Sar power plant, Gilan, Iran. For this purpose stored data include measurements of relative and absolute vibration of shaft bearings in both turbine and compressor sections. Signal processing techniques and mathematical transformations are used for feature extraction, as well as supervised and unsupervised methods for dimensionality reduction. Finally neural networks are employed for classification task and fault detection results for different methods are compared and discussed. Proposed techniques show zero FAR and MAR, when PNN is used with PCA or when MLP or RBF is used with LDA for dimensionality reduction.

Gas turbine shaft unbalance fault detection by using vibration data and neural networks
Mostafa Tajik, Shirin Movasagh, Mahdi Aliyari Shoorehdeli, Iman Yousefi
Robotics and Mechatronics (ICROM), 2015 3rd RSI International Conference on
2015Conference
Abstract:

In this study, seismic attributes have been used to estimate well logs in one of the Iranian petroleum reservoirs. Three static methods have been evaluated: the linear model, the multilayer perceptron (MLP) and the radial basis function (RBF). For linear case, the selection of appropriate attributes was determined by forward selection and for nonlinear one, the selection was based on the genetic algorithm (GA) result. Parameters of nonlinear models were determined by cross-validation and then well logs were estimated. By comparing estimated and actual logs, RBF has the best performance with least training error. Since well logs contain high frequency content, so localized networks such as RBF has better performance than MLP through the study data set.

Petroleum reservoir properties estimation using neural networks
Marzieh Tavasoli, Mahdi Aliyari Shooredeli, Mohammad Ali Nekoui, Majid Fahimi Najm
Fuzzy and Intelligent Systems (CFIS), 2015 4th Iranian Joint Congress on
2015Conference
Abstract:

The walking beam furnace is one of the most prominent process plants often met in an alloy steel production factory and characterised by high non-linearity, strong coupling, time delay, large time-constant and time variation in its parameter set and structure. From another viewpoint, the walking beam furnace is a distributed- parameter process in which the distribution of temperature is not uniform. Hence, this process plant has complicated non-linear dynamic equations that have not worked out yet. In this paper, we propose one-step non-linear predictive model for a real walking beam furnace using non-linear black-box subsystem identification based on locally linear neuro-fuzzy model. Furthermore, a multi-step predictive model with a precise long prediction horizon (i. e., ninety seconds ahead), developed with application of the sequential one-step predictive models, is also presented for the first time. The locally linear model tree which is a progressive tree-based algorithm trains these models. Comparing the performance of the one-step linear neuro-fuzzy model predictive models with their associated models obtained through least squares error solution proves that all operating zones of the walking beam furnace are of non-linear sub-systems. The recorded data from Iran Alloy Steel factory is utilized for identification and evaluation of the proposed neuro-fuzzy predictive models of the walking beam furnace process.

Short-term and long-term thermal prediction of a walking beam furnace using neuro-fuzzy techniques
Hamed Dehghan Banadaki, Hasan Abbasi Nozari, Mahdi Aliyari Shoorehdeli
Thermal Science
2015Journal
Abstract:

This paper presents a variable structure rule-based fuzzy control for trajectory tracking and vibration control of a flexible joint manipulator by using chaotic anti-control. Based on Lyapunov stability theory for variable structure control and fuzzy rules, the nonlinear controller and some generic sufficient conditions for global asymptotic control are attained. The fuzzy rules are directly constructed subject to a Lyapunov function obtained from variable structure surfaces such that the error dynamics of control problem satisfy stability in the Lyapunov sense. Also in this study, the anti-control is applied to reduce the deflection angle of flexible joint system. To achieve this goal, the chaos dynamic must be created in the flexible joint system. So, the flexible joint system has been synchronized to chaotic gyroscope system. In this study, control and anti-control concepts are applied to achieve the high quality performance of flexible joint system. It is tried to design a controller which is capable to satisfy the control and anticontrol aims. The performances of the proposed control are examined in terms of input tracking capability, level of vibration reduction and time response specifications. Finally, the efficacy of the proposed method is validated through experimentation on QUANSER’s flexible-joint manipulator.

Control of Flexible Joint Manipulator via Variable Structure Rule-Based Fuzzy Control and Chaos Anti-Control with Experimental Validation
KANDROODI MOJTABA ROSTAMI, Faezeh Farivar, SHOOREHDELI MAHDI ALIYARI
INTELLIGENT SYSTEMS IN ELECTRICAL ENGINEERING
2014Journal
Abstract:

In this paper, evolutionary algorithms are proposed to compute the optimal parameters of Gaussian Radial Basis Adaptive Backstepping Control (GRBABC) for chaotic systems. Generally, parameters are chosen arbitrarily, so in several cases this choice can be tedious. Also, stability cannot be achieved when the parameters are inappropriately chosen. The optimal design problems are to introduce optimization algorithms like Genetic Algorithms (GA), Particle Swarm Optimization (PSO) in order to find the optimal parameters which minimize a cost function defined as an error quadratic function. These methods are applied to two chaotic systems; Duffing Oscillator and Lü systems. Simulation results verify that our proposed algorithms can achieve enhanced tracking performance regarding similar methods.

Evolutionary Algorithms to Compute the Optimal Parameters of Gaussian Radial Basis Adaptive Backstepping Control for Chaotic Systems
Faezeh Farivar, Mohammad Ali Nekoui, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Universal Journal of Control and Automation
2014Journal
Abstract:

Electro-hydrostatic actuator (EHA) is a kind of hydraulic system in which fluid is routed directly by pump to the actuator. In this study, a novel adaptive fuzzy-PID controller is developed to improve position controlling performance of an EHA. First of all, design and simulation of an EHA by using multidisciplinary modelling method is presented. This model is evaluated by soft validation method. The whole proposed novel control system is composed of a pair of interconnected subsystems, that is, a simple fuzzy-PID controller (SFPID) and a radial basis function neural network (RBFNN) to enhance the tracking performance. The RBFNN fuzzy-PID control (RBFNNF-PID) is applied to EHA. Also, SFPID control, fuzzy-PID control based on extended Kalman filter using grey predictor (FPIDKG) and simple adaptive control (SAC) as significant controls are applied to EHA. The simulation results have shown a significant improvement in transient response and reduction in sum square error (SSE).

Multidisciplinary modelling and position tracking control of an electro-hydrostatic actuator using a novel adaptive fuzzy-PID controller
Mohammad Javad Mirshojaeian Hosseini, Mahdi Aliyari Shoorehdeli
International Journal of Advanced Mechatronic Systems
2014Journal
Abstract:

This study presents the normative knowledge source for the belief space of cultural algorithm(CA) based on an adaptive Radial Basis Function Neural Network (RBFNN). The use of the RBFNN makes it possible to use the previous upper and lower bounds of the normative knowledge to update them and to extract a logical relationship between the previous parameters of the normative knowledge and their new values. The proposed algorithm(N3KCA) is similar to what the human brain does, i.e. to predict the new values of the bounds of normative knowledge based on the previous ones and some knowledge, which it has gained from the previous successive updates. Finally, the proposed cultural algorithm is evaluated on 10 unimodal and multimodal benchmark functions. The algorithm is compared with several other optimization algorithms including previous version of cultural algorithm. In order to have a fair comparison, the number of cost function evaluation is kept the same for all optimization algorithms. The obtained results show that the proposed modification enhances the performance of the CA in terms of convergence speed and global optimality.

Neural Networks for Normative Knowledge Source of Cultural Algorithm
Vahid Seydi Ghomsheh, Mohamad Teshnehlab, Mahdi Aliyari Shoorehdeli, Mojaba Ahmadieh Khanesar
International Journal of Computational Intelligence Systems
2014Journal
Abstract:

Prediction of seasonal influenza epidemics is certainly a forming and effective step towards taking appropriate preventive actions. Improvement on public informing, decreasing the number of infected cases, undesirable effects and deaths due to influenza and also increasing vigilance of Iranian Influenza Surveillance System (IISS), have been practical goals of this research. A forecasting system has been designed and developed using Artificial Neural Networks (ANNs). It is a novel research as a novel dataset has been exploited. The data are categorized in two groups of climatic parameters (temperature, humidity, precipitation, wind speed & sea level pressure) and number of patients (number of total referrals and number of patients with Influenza-Like Illnesses (ILI)). In order to evaluate the model performance, different cost functions are defined and results indicate that the model provides the possibility of a satisfactory forecasting and is practically helpful to achieve the objectives already claimed.

Prediction of seasonal influenza epidemics in Tehran using artificial neural networks
Fatemeh Saberian, Ali Zamani, Mohammad Mahdi Gooya, Payman Hemmati, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Electrical Engineering (ICEE), 2014 22nd Iranian Conference on
2014Conference
Abstract:

A real-time dynamic hardware-in-loop (HIL) simulator of an RTX real-time subsystem (RTSS) was developed by using LabVIEW (G language). The main idea of this work was to determine the feasibility and accuracy of widely available and highly competitive commercial products, such as personal computers on an RTSS, as an alternative to conventional prohibitive real-time simulators in dynamic studies of power systems. The implemented system is a self-contained heavy-duty gas turbine, governor, synchronous 200-MVA, 15.75-kV machine and a simplified electrical network. The HIL simulator was customized to interact with a 1518-kW static exciter. The role of this HIL simulator is to provide real-time digital and analog signals for static exciter systems (SES) and to simulate the mechanical and electrical components in a closed-loop, fixed-step solver applied by a well-known numerical solution method. This sophisticated yet exceptionally economic HIL simulator provides engineers with a safe environment to analyze the dynamic performance of static exciters and investigate their natural restraints and functionalities. It also provides a safe environment to analyze some naturally destructive tests.

Real-time dynamic HIL simulator of gas turbine, governor, generator and grid for static excitation of a 200-MVA synchronous generator
Mohamad Esmaeil Iranian, Iman Yousefi, Mahdi Aliyari Shoorehdeli
Simulation
2014Journal
Abstract:

Modern systems are required to guarantee a high degree of safety and self-diagnostics capabilities. This paper investigates the problem of state fault diagnosis in nonlinear systems with modeling uncertainties. In contrast with common literature, the fault diagnosis scheme is proposed in discrete time domain. This property relaxes the risk of stability and performance degradation in deriving discrete equivalent of continuous methods. An estimator is designed in order to generate residual signal by utilizing a proper nonlinear state transformation. A robust compensator term is implemented in estimator to decrease effect of modeling uncertainties and approximation error on residual signal. When the residual signal is exceeded detection threshold, an on-line fault approximator is turned on and trained by appropriate parameter update law. An extra term is considered in update rule to overcome the need of persistency of excitation (PE). The implement of all robust compensator term, PE relaxing term and proper parameter adaption law improve the accuracy of fault reconstruction. The result would be obviously vital in tolerant and time-life prediction stages after fault diagnosis.

Robust state fault diagnosis in nonlinear discretetime systems with modelling uncertainties; using an automated intelligent methodology
Leila Mahmoodi, Mahdi Aliyari Shoorehdeli
Smart Grid Conference (SGC), 2014
2014Conference
Abstract:

In this paper, Fault Detection and Isolation (FDI) is studied for the rotary kiln of Saveh White Cement Company. To do so, K-means algorithm as a crisp clustering, Fuzzy C-Means (FCM), and Gustafson-Kessel (GK) algorithms as fuzzy clustering are used. In those, for finding number of clusters, Cluster Validity Indices (CVI) are applied. Principal Component Analysis (PCA) mapped the clusters into two dimensional spaces. Fault detection and isolation performance are evaluated by three criteria namely sensitivity, specificity, and confusion matrix. The results reveal that GK fuzzy algorithm provides better performance on detection and isolation of fault in this industrial plant.

Fault Detection and Isolation of a cement rotary kiln using fuzzy clustering algorithm
Nayereh Talebnezhad, Alireza Fatehi, Mahdi Aliyari Shoorehdeli
Fuzzy Systems (IFSC), 2013 13th Iranian Conference on
2013Conference
Abstract:

In this study, fault tolerant control for a Rotary Inverted Pendulum (RIP) has been improved by using chaos synchronization with adding a chaotic signal as a reference. Rotary inverted pendulum is a nonlinear, under-actuated, unstable and non-minimum-phase system. The proposed control consists of a state-feedback (LQR) and a fuzzy-PID control. The state- feedback control is used to stabilize system near the operating point, and the fuzzy-PID is used to track the chaos signal. PID controller gains adjust by fuzzy rule. The designed controller is implemented on a Quanser laboratory system.

Fault tolerant improvement with chaos synchronization using Fuzzy-PID control
Farhad Ghorbani, Mahdi Aliyari Shooredeli, Mohammad Teshnehlab
Fuzzy Systems (IFSC), 2013 13th Iranian Conference on
2013Conference
Abstract:

MiRNAs play an essential role in the networks of gene regulation by inhibiting the translation of target mRNAs. Several computational approaches have been proposed for the prediction of miRNA target-genes. Reports reveal a large fraction of under-predicted or falsely predicted target genes. Thus, there is an imperative need to develop a computational method by which the target mRNAs of existing miRNAs can be correctly identified. In this study, combined pattern recognition neural network (PRNN) and principle component analysis (PCA) architecture has been proposed in order to model the complicated relationship between miRNAs and their target mRNAs in humans. The results of several types of intelligent classifiers and our proposed model were compared, showing that our algorithm outperformed them with higher sensitivity and specificity. Using the recent release of the mirBase database to find potential targets of miRNAs, this model incorporated twelve structural, thermodynamic and positional features of miRNA:mRNA binding sites to select target candidates.

HomoTarget: a new algorithm for prediction of microRNA targets in Homo sapiens
Hamed Ahmadi, Ali Ahmadi, Sadegh Azimzadeh-Jamalkandi, Mahdi Aliyari Shoorehdeli, Ali Salehzadeh-Yazdi, Gholamreza Bidkhori, Ali Masoudi-Nejad
Genomics
2013Journal
Abstract:

In this study, first by using the collected real data from a 10000 cubic-meter Qazvin-kowsar water supply reservoir is modeled by nonlinear output error (NOE) structure, then a neural nonlinear controller based on the MLP neural network according to created model is designed in order to control the tank water level. The operation of the proposed controller is compared by a PID controller which its coefficients is optimized by genetic algorithm. Results of the simulation indicates that the neural nonlinear controller has a better function than the PID controller, and also this controller is able to control the level water of the tank appropriately regardless the consumer profile in all conditions even in consumer picks.

Identification and control of water supply reservoirs by using neural networks
Mahdi Keshavarz Ghasemi, Mahdi Aliyari Shoorehdeli
Fuzzy Systems (IFSC), 2013 13th Iranian Conference on
2013Conference
Abstract:

This brief presents a X–Y pedestal using the feedback error learning (FEL) controller with adaptive neural network for low earth orbit (LEO) satellite tracking applications. The aim of the FEL is to derive the inverse dynamic model of the X–Y pedestal. In this brief, the kinematics of X–Y pedestal is obtained. To minimize or eliminate the backlash between gears, an antibacklash gearing system with dual-drive technique is used. The X–Y pedestal is implemented and the experimental results are obtained. They verify the obtained kinematics of the X–Y pedestal, its ability to minimize backlash, and the reduction of the tracking error for LEO satellite tracking in the typical NOAA19 weather satellite. Finally, the experimental results are plotted.

Implementation and Control of X–Y Pedestal Using Dual-Drive Technique and Feedback Error Learning for LEO Satellite Tracking
Taheri, A.,M.Aliyari Sh., H. Bahrami, M.H. Fatehi
Control Systems Technology, IEEE Transactions on
2013Journal
Abstract:

In this study it is attempted to describe the structure and procedure of training for the Interval Type-2 Fuzzy Logic inference System completely. To achieve this goal Adaptive Network- based Fuzzy Inference System (ANFIS) structure has been generalized to interval type-2 fuzzy, also all of the relations to describe inference structure and all of the necessary differentiation to adjust parameters with Gradient descent and Levenberg-Marquardt method has been brought. Described structure has been used to forecast Mackey-Glass chaotic time- series that polluted with additive uncertain domain noise. Using mentioned procedure for parameters adjustment achieved acceptable results.

Interval type-2 adaptive network-based fuzzy inference system (ANFIS) with Type-2 non-singleton fuzzification
Hossein MonirVaghefi, Mohsen Rafiee Sandgani, Mahdi Aliyari Shoorehdeli
Fuzzy Systems (IFSC), 2013 13th Iranian Conference on
2013Conference
Abstract:

This study proposes a novel chaotic anti-control for flexible joint system. The proposed controller is composed of a Lyapunov rule-based fuzzy control and chaotic anti-control for target tracking of the flexible joint manipulator. Chaotic signal is used to study the effect of anti-control to reduce the deflection of flexible joint system and control signal energy. For this purposes the flexible joint has been synchronized with chaotic Lorenz system. In this study on of the Lorenz parameters is changed to analysis the effect of chaotic signals. The results of the proposed approach shows in terms of level of vibration reduction and energy consumption of control signal, we could find an optimum point based on value of Lorenz system parameter. Finally, the efficacy of the proposed method and results of existence of different nonlinearity behavior is validated through experiments on QUANSER's flexible-joint manipulator.

Lyapunov rule-based fuzzy control and chaotic anti-control for flexible joint system and analysis of chaotic signal existence effectiveness with experimental validation
Mohsen Rafiee Sandgani, Mahdi Aliyari Shoorehdeli
Fuzzy Systems (IFSC), 2013 13th Iranian Conference on
2013Conference
Abstract:

Neural network based controller is used for controlling a mobile robot system. Feedback error learning (FEL) can be regarded as a hybrid control to guarantee stability of control approach. This paper presents simulation of a mobile robot system controlled by a FEL neural network and PD controllers. This feedback error-learning controller benefits from both classic and adaptive controller properties. The simulation results demonstrate that this method is more feasible and effective for mobile robot system control.

Mobile robot control based on neural network and feedback error learning approach
H Zarabadipour, Z Yaghoubi, M Aliyari Shoorehdeli
strategies
2013Journal
Abstract:

In this study interval type-2 fuzzy systems with non-singleton type-2 fuzzifire are used for identification and modeling nonlinear systems having noise with changing domain for fault detection purpose. The main idea in this fault detection method is to serve an upper bound and a lower bound as a confidence bound for system output that obtained from the interval type-2 fuzzy system. If we haven't precise information about mean and variance of noise, then non-singleton type-2 fuzzifire is usable. This fuzzifire improves performance of fault detection confidence bound. In the end of this paper a well-known benchmark two-tank system has been used for representing the advantages of proposed fault detection method.

Model-based fault detection of a nonlinear system using interval type-2 fuzzy systems with non-singleton type-2 fuzzification
Hossein Monirvaghefi, Mahdi Aliyari Shoorehdeli
Control, Instrumentation, and Automation (ICCIA), 2013 3rd International Conference on
2013Conference
Abstract:

In this paper, modeling, identification and control of a real 162MW heavy duty industrial gas turbine is taken into account. This work is aimed to introduce a simple and comprehensive model to test various controllers. Rowen's model is used to present the mechanical behavior of the gas turbine, while the identification of it is done using a feedforward neural network. The control rules of the turbine are applied on both models and a comparison of the results is also presented.

Modeling, identification and control of a heavy duty industrial gas turbine
Iman Yousefi, Mostafa Yari, Mahdi Aliyari Shoorehdeli
Mechatronics and Automation (ICMA), 2013 IEEE International Conference on
2013Conference
Abstract:

This paper proposes the modified projective synchronization method for unknown chaotic dissipative gyroscope systems via Gaussian radial basis adaptive variable structure control. Because of the nonlinear terms of the dissipative gyroscope system, the system exhibits chaotic motions. As chaotic signals are usually broadband and noise-like, synchronized chaotic systems can be used as cipher generators for secure communication. Obviously the importance of obtaining these objectives is specified when the dynamics of the gyroscope system are unknown. In this paper, using the neural variable structure control technique, control laws are established, which guarantees the modified projective synchronization of an unknown chaotic dissipative gyroscope system. Switching surfaces are adopted to ensure the stability of the error dynamics in variable structure control. In the neural variable structure control, Gaussian radial basis functions are utilized online to estimate the system dynamic functions. Also, the adaptation laws of the online estimators are derived in the sense of Lyapunov function. Thus, the unknown chaotic gyroscope system can be guaranteed to be asymptotically stable. Also, the synchronization objectives have been achieved. The proposed method allows us to arbitrarily adjust the desired scaling by controlling the slave system. It is not necessary to calculate the Lyapunov exponents and the eigenvalues of the Jacobian matrix, which makes it simple and convenient. Also, it is a systematic procedure for modified projective synchronization of chaotic systems and it can be applied to a variety of chaotic systems no matter whether it contains external excitation or not. The designed control system is robust versus model uncertainty. Numerical simulations are presented to verify the proposed synchronization method.

Modified projective synchronization of unknown chaotic dissipative gyroscope systems via Gaussian radial basis adaptive variable structure control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab, Mohammad Ali Nekoui
Journal of Vibration and Control
2013Journal
Abstract:

This paper deals with the issue of position control of an Electro-Hydrostatic Actuator (EHA) using an adaptive PID controller based on neurofuzzy network. In this relation, the design and simulation of an electro-hydrostatic actuation system referred to as EHA using multidisciplinary modeling method is presented. In recent years, fuzzy-PID controller is one of the main controllers that apply to the EHA systems. To improve the response of this controller, another control technique is needed to combine with the fuzzy-PID, and also, training some parameters of fuzzy-PID technique is a solution. The whole of new controller is composed of pair of interconnected subsystems, that is, an RBF network and conventional fuzzy-PID controller to enhance the tracking performance. Results show a significant improvement in transient response is achieved in comparison with a conventional fuzzy-PID control.

Multidisciplinary modeling and position control of an electro-hydrostatic actuation system using an adaptive PID controller based on neurofuzzy network
Mohammad Javad Mirshojaeian Hosseini, Soheil Alidoosti, Mahdi Aliyari Shoorehdeli
Control, Instrumentation, and Automation (ICCIA), 2013 3rd International Conference on
2013Conference
Abstract:

Many successful methods in various vision tasks rely on statistical analysis of visual patterns. However, we are interested in covering the gap between the underlying mathematical representation of the visual patterns and their statistics. With this general trend, in this paper a relationship between phase structure of a class of patterns and their moments after and before filtering have been considered. First, a general formula between the phase structure and moments of the images is obtained. Second, a theorem is developed that states under which conditions two visual patterns with the same frequencies but different phases have the same moments up to a certain moment. Finally, a theorem is developed that explains, given a set of filters, under which conditions two visual patterns with both different frequencies and different phases have the same subband statistics.

Patterns with different phases but same statistics
Peyman Sheikholharam Mashhadi, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
JOSA A
2013Journal
Abstract:

In this study, a new adaptive controller is proposed for position control of pneumatic systems. Difficulties associated with the mathematical model of the system in addition to the instability caused by Pulse Width Modulation (PWM) in the learning-based controllers using gradient descent, motivate the development of a new approach for PWM pneumatics. In this study, two modified Feedback Error Learning (FEL) methods are suggested and the their effectiveness are validated by experimental tracking data. The first one is a combination of PD (Proportionalâ Derivative) and RBF (Radial Basis Function) and in the second one RBF is replaced by ANFIS (Adaptive Neuro-Fuzzy Inference System). The robustness to varying mass is also examined. The experimental results show that the proposed algorithms, especially with ANFIS, are able to give good performance regardless of any uncertainties.

Position Control of a Pulse Width Modulated Pneumatic Systems: an Experimental Comparison
Mahdi Aliyari Shoorehdeli, Farid Najafi, Sahar Jafari
International Journal of Robotics, Theory and Applications
2013Journal
Abstract:

MR-based methods have acceded an important role for the clinical detection and diagnosis of breast cancer. Dynamic contrast-enhanced MRI of the breast has become a robust and successful method, especially for the diagnosis of high-risk cases due to its higher sensitivity compared to X-ray mammography. In the clinical setting, the ANN has been widely applied in breast cancer diagnosis using a subjective impression of different features based on defined criteria. In this study, several neural networks classifiers like MLP, PNN, GRNN, and RBF has been presented on a total of 112 histopathologically verified breast lesions to classify into benign and malignant groups. Also, support vector machine has been considered as classifier. Before applying classification methods, feature selection has been utilized to choose the significant features for classification. Finally, to improve the performance of classification, three classifiers that have the best results among all applied methods have been combined together that they been named as multi-classifier system. For each lesion, final detection as malignant or benign has been evaluated, when the same results have been achieved from two classifiers of multi-classifier system. Tables of results show that the proposed methods are correctly capable to feature selection and improve classification of breast cancer.

Specificity enhancement in classification of breast MRI lesion based on multi-classifier
Farzaneh Keyvanfard, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab, Ke Nie, Min-Ying Su
Neural Computing and Applications
2013Journal
Abstract:

This study applies technique PDC (parallel distributed compensation) for speed control of a Digital Servo System. PDC method is based on nonlinear Takagi-Sugeno (TS) fuzzy model. Also in this study Neural Adaptive is used for velocity control and identification of a Digital Servo System. It is shown that these techniques can be successfully used to stabilize any chosen operating point of the system. All derived results are validated by experimental and computer simulation. The controllers which introduced have big range for control the system. We compare PDC controller with Neural Adaptive controller results and PID controller.

Speed control of a Digital Servo System using parallel distributed compensation controller and Neural Adaptive controller
Zohreh Alzahra Sanai Dashti, Milad Gholami, Mohammad Jafari, M Aliyari Shoorehdeli, M Teshnehlab
Fuzzy Systems (IFSC), 2013 13th Iranian Conference on
2013Conference
Abstract:

In this paper, a new steganalysis method based on Cellular Automata Transform (CAT) is presented. CAT is used for feature extraction from stego and clean images. For that purpose, three levels CAT is applied on images and 12 sub-bands are generated for feature extraction. With adding the original image, 13 sub-bands are be used in feature extraction phase. In the next step, three moments of characteristic function (CF) are used as feature vector for every image (stego or clean image). At the end, Neural Network (NN) is applied as classifier. This supervised learning method uses these features for classifying the input image into either stego-image or clean-image. The performance of this algorithm is verified using some test samples. The results of our empirical tests show that detection accuracy of our method reaches to 93% for breaking MB2 and 91% for breaking LSB. Therefore the proposed method is a blind steganalysis method that can be used for broking some steganography methods.

Steganalysis algorithm based on Cellular Automata Transform and Neural Network
Soodeh Bakhshandeh, Fateme Bakhshande, Mahdi Aliyari
Information Security and Cryptology (ISCISC), 2013 10th International ISC Conference on
2013Conference
Abstract:

This brief proposes modified projective synchronization (MPS) methods for underactuated unknown heavy symmetric chaotic gyroscope systems via optimal Gaussian radial basis adaptive variable structure control. Chaotic gyroscope systems are considered as underactuated systems where a control input is designed to synchronize the two degree of freedoms interactions. Until now, no investigation of this subject with one control input has been presented. The importance of obtaining synchronization objectives is specified when the dynamics of gyroscope system are unknown. In this brief, using the neural variable structure control technique, a control law is established that guarantees the MPS of underactuated unknown chaotic gyros. In the neural variable structure control, Gaussian radial basis functions are utilized to estimate online the system dynamic functions. Adaptation laws of the online estimator are derived in the sense of the Lyapunov function. Moreover, online and offline optimizers are applied to optimize the energy of the control signal. The proposed solution is generalized to chaos control of the mentioned gyroscopes. Numerical simulations are presented to verify the proposed synchronization methods.

Synchronization of underactuated unknown heavy symmetric chaotic gyroscopes via optimal Gaussian radial basis adaptive variable structure control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
IEEE Transactions on control systems technology
2013Journal
Abstract:

This paper shows a new fuzzy system was improved using genetic algorithm to handle fuzzy inference system as a function approximator and time series predictor. The system was developed generality that trained with genetic algorithms (GAs) corresponding to special problem and would be evaluated with different number of rules and membership functions. Then, compare the efficacy of variation of these two parameters in behavior of the system and show the method that achieves an efficient structure in both of them. Also, the proposed GA-Fuzzy inference system successfully predicts a benchmark problem and approximates an introduced function and results have been shown.

The Scrutiny of Variation in the Number of Fuzzy Rules and Membership Functions in a New Genetic-Fuzzy System in Approximation and Prediction Problems
Vahideh Keikha, Hayat Khoobipour, Mahdi Aliyari Shoorehdeli, Hassan Rezaei
International Journal of Information and Electronics Engineering
2013Journal
Abstract:

Type-1 fuzzy sets cannot fully handle the uncertainties. To overcome the problem, type-2 fuzzy sets have been proposed. The novelty of this paper is using interval type-2 fuzzy logic controller (IT2FLC) to control a flexible-joint robot with voltage control strategy. In order to take into account the whole robotic system including the dynamics of actuators and the robot manipulator, the voltages of motors are used as inputs of the system. To highlight the capabilities of the control system, a flexible joint robot which is highly nonlinear, heavily coupled and uncertain is used. In addition, to improve the control performance, the parameters of the primary membership functions of IT2FLC are optimized using particle swarm optimization (PSO). A comparative study between the proposed IT2FLC and type-1 fuzzy logic controller (T1FLC) is presented to better assess their respective performance in presence of external disturbance and unmodelled dynamics. Stability analysis is presented and the effectiveness of the proposed control approach is demonstrated by simulations using a two-link flexible-joint robot driven by permanent magnet direct current motors. Simulation results show the superiority of the IT2FLC over the T1FLC in terms of accuracy, robustness and interpretability.

Type-2 fuzzy control for a flexible-joint robot using voltage control strategy
Majid Moradi Zirkohi, Mohammad Mehdi Fateh, Mahdi Aliyari Shoorehdeli
International Journal of Automation and Computing
2013Journal
Abstract:

This paper presents the identification of V94.2 gas turbine. This turbine is built by Siemens. It has 162.1 MW nominal power and 50 Hz nominal frequency and is located at Kermanshah power plant, Kermanshah city of Iran. The stored data from turbine include fuel pressure valve angle and IGV1 angle as inputs and compressor output pressure, compressor output temperature, fuel pressure, turbine output power and turbine output temperature as outputs. To simplify identification process, the system turns into MISO2 systems to the number of outputs, and then correlation analysis is used to examine the dependence of the outputs to each input and other outputs. For turbine identification, dynamic linear models are estimated and then Feedforward neural network with one hidden layer is trained. The result shows dynamic linear models have poor performance in comparison with Feedforward neural network with one hidden layer. The neural network is able to identify a predictor model with fitness over 96% for outputs of V94.2 gas turbine.

V94. 2 gas turbine identification using neural network
Mostafa Yari, Mahdi Aliyari Shoorehdeli, Iman Yousefi
Robotics and Mechatronics (ICRoM), 2013 First RSI/ISM International Conference on
2013Conference
Abstract:

Despite active research and significant progress in the last three decades on control of human eye movements, it remains challenging issue due to its applications in prosthetic eyes and robotics. Till now, no considerable investigation of this subject is presented in the interdisciplinary sciences. The goal of this paper is to present a distinguished survey of existing literature on the intelligent control of the human eye movements system applied in a huggable pet-type robot as a biomechatronic system. In this study, the basic knowledge of human eye movements control is explained to show how the neural networks in the brainstem control the human eye movements. The geometry and model of human eye movements system are investigated and this system is considered as a nonlinear control system. The specified model may only be an academic exercise. It can have scientific importance in understanding of the human movement system in general. Also, it can be useful for robotics. Intelligent methods such as artificial neural networks and fuzzy neural networks are proposed to control the human eye movements and numerical simulations are presented. It is discussed that the intelligent controls applied to control of human eye movements system are emulated from the neural controls in biological system.

An interdisciplinary overview and intelligent control of human prosthetic eye movements system for the emotional support by a huggable pet-type robot from a biomechatronical viewpoint
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Journal of the Franklin Institute
2012Journal
Abstract:

This paper addresses the experimental identification of a servo actuator which is used in many industrial applications. Because the system consisted of electrical and mechanical components, the behavior of the system was nonlinear. In addition, the under load behavior of this servo was different. The load torque was considered as the input and a two input-one output model was presented for this servo actuator. Special was given in order to present a simple and applicable model for this servo actuator. For identification of this servo actuator, classic and intelligent methods have been used. ARMAX model as a classic model and MLP and LOLIMOT networks as intelligent models were selected for this purpose and their results have been discussed. The comparisons between these methods show that the intelligent methods have a better accuracy than classical method, but they have more complexity in the implementation. These models can be applied as references for characterizing different designs and future control strategies.

An under load servo actuator identification and comparison between the results of different methods
M Maboodi, MH Ashtari Larki, M Aliyari Shoorehdeli
Iranian Journal of Electrical & Electronic Engineering
2012Journal
Abstract:

Breast cancer is the cause of the most common cancer death in women. Early detection of the breast cancer is an effective method to reduce mortality. Fuzzy Neural Networks (FNN) comprises an integration of the merits of neural and fuzzy approaches, enabling one to build more intelligent decision-making systems. But increasing the number of inputs causes exponential growth in the number of parameters in Fuzzy Neural Networks (FNN) and computational complexity increases accordingly. This phenomenon is named as “curse of dimensionality”. The Hierarchical Fuzzy Neural Network (HFNN) and the Fuzzy Gaussian Potential Neural Network (FGPNN) are utilized to deal this problem. In this study, the HFNN and FGPNN by using new training algorithm, are applied to the Wisconsin Breast Cancer Database to classify breast cancer into two groups; benign and malignant lesions. The HFNN consists of hierarchically connected low-dimensional fuzzy neural networks. It can use fewer rules and parameters to model nonlinear system. Moreover, the FGPNN consists of Gaussian Potential Function (GPF) used in the antecedent as the membership function. When the number of inputs increases in FGPNN, the number of fuzzy rules does not increase. The performance of HFNN and FGPNN are evaluated and compared with FNN. Simulation results show the effectiveness of these methods even with less rules and parameters in performance result. These methods maintain the accuracy of original fuzzy neural system and have high interpretability by human in diagnosis of breast cancer.

Breast cancer classification based on advanced multi dimensional fuzzy neural network
Somayeh Naghibi, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
Journal of medical systems
2012Journal
Abstract:

This paper proposes the chaos control and the generalized projective synchronization methods for heavy symmetric gyroscope systems via Gaussian radial basis adaptive variable structure control. Because of the nonlinear terms of the gyroscope system, the system exhibits chaotic motions. Occasionally, the extreme sensitivity to initial states in a system operating in chaotic mode can be very destructive to the system because of unpredictable behavior. In order to improve the performance of a dynamic system or avoid the chaotic phenomena, it is necessary to control a chaotic system with a periodic motion beneficial for working with a particular condition. As chaotic signals are usually broadband and noise like, synchronized chaotic systems can be used as cipher generators for secure communication. This paper presents chaos synchronization of two identical chaotic motions of symmetric gyroscopes. In this paper, the switching surfaces are adopted to ensure the stability of the error dynamics in variable structure control. Using the neural variable structure control technique, control laws are established which guarantees the chaos control and the generalized projective synchronization of unknown gyroscope systems. In the neural variable structure control, Gaussian radial basis functions are utilized to on-line estimate the system dynamic functions. Also, the adaptation laws of the on-line estimator are derived in the sense of Lyapunov function. Thus, the unknown gyro systems can be guaranteed to be asymptotically stable. Also, the proposed method can achieve the control objectives. Numerical simulations are presented to verify the proposed control and synchronization methods. Finally, the effectiveness of the proposed methods is discussed.

Chaos control and generalized projective synchronization of heavy symmetric chaotic gyroscope systems via Gaussian radial basis adaptive variable structure control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
Chaos, Solitons & Fractals
2012Journal
Abstract:

This paper proposes the chaos control and the modified projective synchronization methods for unknown heavy symmetric chaotic gyroscope systems via Gaussian radial basis adaptive backstepping control. Because of the nonlinear terms of the gyroscope system, the system exhibits chaotic motions. Occasionally, the extreme sensitivity to initial states in a system operating in chaotic mode can be very destructive to the system because of unpredictable behavior. In order to improve the performance of a dynamic system or avoid the chaotic phenomena, it is necessary to control a chaotic system with a regular or periodic motion beneficial for working with a particular condition. As chaotic signals are usually broadband and noise-like, synchronized chaotic systems can be used as cipher generators for secure communication. Obviously, the importance of obtaining these objectives is specified when the dynamics of gyroscope system are unknown. In this paper, using the neural backstepping control technique, control laws are established which guarantees the chaos control and the modified projective synchronization of unknown chaotic gyroscope system. In the neural backstepping control, Gaussian radial basis functions are utilized to on-line estimate the system dynamic functions. Also, the adaptation laws of the on-line estimators are derived in the sense of Lyapunov function. Thus, the unknown chaotic gyroscope system can be guaranteed to be asymptotically stable. Also, the control objectives have been achieved. The proposed method allows us to arbitrarily adjust the desired scaling by controlling the slave system. It is not necessary to calculate the Lyapunov exponents and the eigenvalues of the Jacobian matrix, which makes it simple and convenient. Also, it is a systematic procedure for modified projective synchronization of chaotic systems and it can be applied to a variety of chaotic systems no matter whether it contains external excitation or not. Notice that it needs only one controller to realize modified projective synchronization no matter how much dimensions the chaotic system contains and the controller is easy to be implemented. It seems that the proposed method can be useful for practical applications of chaotic gyroscope systems in the future. Numerical simulations are presented to verify the proposed control and synchronization methods.

Chaos control and modified projective synchronization of unknown heavy symmetric chaotic gyroscope systems via Gaussian radial basis adaptive backstepping control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
Nonlinear Dynamics
2012Journal
Abstract:

A novel structure of fuzzy logic controller is presented for trajectory tracking and vibration control of a flexible joint manipulator. The rule base of fuzzy controller is divided into two sections. Each section includes two variables. The variables of first section are the error of tip angular position and the error of deflection angle, while the variables of second section are derivatives of mentioned errors. Using these structures, it would be possible to reduce the number of rules. Advantages of proposed fuzzy logic are low computational complexity, high interpretability of rules, and convenience in fuzzy controller. Implementing of the fuzzy logic controller on Quanser flexible joint reveals efficiency of proposed controller. To show the efficiency of this method, the results are compared with LQR method. In this paper, experimental validation of proposed method is presented.

Control of flexible joint manipulator via reduced rule-based fuzzy control with experimental validation
Mojtaba Rostami Kandroodi, Mohammad Mansouri, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
ISRN Artificial Intelligence
2012Journal
Abstract:

In a cement factory, a rotary kiln is the most complex component and it plays a key role in the quality and quantity of the final product. This system involves complex nonlinear dynamic equations that have not been completely worked out yet. In conventional modeling procedures, a large number of the involved parameters are crossed out and an approximation model is presented instead. Therefore, the performance of the obtained model is very important and an inaccurate model may cause many problems in the design of a controller. This study presents a Takagi-Sugeno (TS)-type fuzzy system called a wavelet projection fuzzy inference system (WPFIS) in which a dimension reduction section is used at the input stage of the fuzzy system. In order to clarify the structure of the extracted features, structural learning with forgetting (SLF) based on Minkowski norms is proposed. In addition, gradient descent (GD) was used as a training algorithm. The results show that the proposed method has higher performance in comparison with conventional models. The data collected from Saveh White Cement Company were used in our simulations.

Design of a prediction model for cement rotary kiln using wavelet projection fuzzy inference system
A Sharifi, M Aliyari Shoorehdeli, Mohammad Teshnehlab
Cybernetics and Systems
2012Journal
Abstract:

This paper presents energy reduction with anticontrol of chaos for nonholonomic mobile robot system. Anticontrol of chaos is also called chaotification, meaning to chaotify an originally non-chaotic system, and in this paper error of mobile robot system has been synchronized with chaotic gyroscope for reducing energy and increasing performance. The benefits of chaos synchronization with mechanical systems have led us to an innovation in this paper. The main purpose is that the control system in the presence of chaos work with lower control cost and control effort has been reduced. For comparison of proposed method, the feedback linearization controller has also been designed for mobile robot with noise. Finally, the efficacies of the proposed method have been illustrated by simulations, energy of control signals has been calculated, and effect of Alpha (: a constant coefficient is used beside of chaotic system) variations on the energy of control signals has been checked.

Energy reduction with anticontrol of chaos for nonholonomic mobile robot system
Zahra Yaghoubi, Hassan Zarabadipour, Mahdi Aliyari Shoorehdeli
Abstract and Applied Analysis
2012Journal
Abstract:

In this paper Fault Detection and Isolation (FDI) is shown as a pattern classification problem which can be solved using clustering techniques. Gath-Geva clustering (GGC) is exploited as optimal form by a performance assessment rule for fault detection, while multistage Gath-Geva clustering is employed for the intent of fault isolation. Furthermore since Visbreaker unit is a large scale process, a novel hybrid method on the basis of Principle Component Analysis and Genetic Algorithm optimization was also proposed in order to cope with the curse of dimensionality and complexity of computation problems. There are two main percentile criteria for validation of fault detection namely specificity and sensitivity. Evaluation of fault isolation has been depicted in confusion matrix. For analysis and visualization of the correlated high dimensional data, PCA maps the data point into lower dimensional space. The proposed FDI approaches have been evaluated through experimental Visbreaker process unit data collected in oil refinery.

Fault detection and isolation of Visbreaker unit in oil refinery using multistage Gath-Geva clustering
Mohammad Mokhtare, Somayeh Hekmati Vahed, Mahdi Aliyari Shoorehdeli, Alireza Fatehi
Electrical Engineering (ICEE), 2012 20th Iranian Conference on
2012Conference
Abstract:

This paper describes hybrid multivariate methods: Fisher's Discriminant Analysis and Principal Component Analysis improved by Genetic Algorithm. These methods are good techniques that have been used to detect faults during the operation of industrial processes. In this study, score and residual space of modified PCA and modified FDA are applied to the Tennessee Eastman Process simulator and show that modified PCA and modified FDA are more proficient than PCA and FDA for detecting faults.

Fault Detection in Tennessee Eastman Process Using Fisher’s Discriminant Analysis and Principal Component Analysis Modified by Genetic Algorithm
Mostafa Noruzi Nashalji, Seyed Mohammad Razeghi, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Applied Mechanics and Materials
2012Journal
Abstract:

Development of a fault detection scheme for nonlinear systems is often difficult due to complexity of the system. In this study a new method, based on parity relations for linear systems, is proposed to detect faults in nonlinear systems that can be modeled by Takagi- Sugeno (TS) fuzzy system. This method is an intuitive generalization of parity relations, because TS fuzzy system uses local linear models. Results of simulation and implementation on a rotary inverted pendulum show that faults can be detected very well.

Fault Detection of Nonlinear Systems Based on Takagi-Sugeno Fuzzy Models by Parity Relations
Majid Ghaniee Zarch, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
A Review of the State of the Art of Modulation Techniques and Control Strategies for Matrix Converters
2012Journal
Abstract:

This study presents fault tolerant control of inverted pendulum via on-line fuzzy backstepping and anti-control of chaos. The inverted pendulum is used frequently in robotic applications and can be found in different forms. Based on Lyapunov stability theory for backstepping design, the nonlinear controller and some generic sufficient conditions for asymptotic control are attained. Also in this study, anti-control of chaos is applied to increase the fault tolerant of inverted pendulum. To achieve this goal, the chaos dynamic must be created in the inverted pendulum system. So, the inverted pendulum system has been synchronized to chaotic gyroscope system. In this study, control and anti-control concepts are applied to achieve the high quality performance of inverted pendulum system. The performances of the proposed control are examined in terms of fault tolerant capability. Finally, the efficacies of the proposed methods are illustrated by simulations.

Fault tolerant control of mechatronics system based on hybrid control
Atefeh Saedian, Hassan Zarabadipour, Mahdi Aliyari Shoorehdeli, Faezeh Farivar
International Journal of Physical Sciences
2012Journal
Abstract:

In this paper, fault tolerant synchronization of chaotic gyroscope systems versus external disturbances via Lyapunov rule-based fuzzy control is investigated. Taking the general nature of faults in the slave system into account, a new synchronization scheme, namely, fault tolerant synchronization, is proposed, by which the synchronization can be achieved no matter whether the faults and disturbances occur or not. By making use of a slave observer and a Lyapunov rule-based fuzzy control, fault tolerant synchronization can be achieved. Two techniques are considered as control methods: classic Lyapunov-based control and Lyapunov rule-based fuzzy control. On the basis of Lyapunov stability theory and fuzzy rules, the nonlinear controller and some generic sufficient conditions for global asymptotic synchronization are obtained. The fuzzy rules are directly constructed subject to a common Lyapunov function such that the error dynamics of two identical chaotic motions of symmetric gyros satisfy stability in the Lyapunov sense. Two proposed methods are compared. The Lyapunov rule-based fuzzy control can compensate for the actuator faults and disturbances occurring in the slave system. Numerical simulation results demonstrate the validity and feasibility of the proposed method for fault tolerant synchronization.

Fault tolerant synchronization of chaotic heavy symmetric gyroscope systems versus external disturbances via Lyapunov rule-based fuzzy control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli
ISA transactions
2012Journal
Abstract:

Rotary kiln is the central and the most complex component of cement production process. It is used to convert calcineous raw meal into cement clinkers, which plays a key role in quality and quantity of the final produced cement. This system has complex nonlinear dynamic equations that have not been completely worked out yet. In conventional modeling procedure, a large number of the involved parameters are crossed out and an approximation model is presented instead. Therefore, the performance of the obtained model is very important and an inaccurate model may cause many problems for designing a controller. This study presents hierarchical wavelet TS-type fuzzy inference system (HWFIS) for identification of cement rotary kiln. In the proposed method, wavelet fuzzy inference system (WFIS) with two input variables is used as sub-model in a hierarchical structure and gradient descent (GD) algorithm is chosen for training parameters of antecedent and conclusion parts of sub-models. The results show that the proposed method has higher performance in comparison with the other models. The data collected from Saveh White Cement Company is used in our simulations.

Identification of cement rotary kiln using hierarchical wavelet fuzzy inference system
A Sharifi, M Aliyari Shoorehdeli, M Teshnehlab
Journal of the Franklin Institute
2012Journal
Abstract:

Differential pipe sticking (DPS) is one of the most conventional and serious problems in drilling operations that imposes some extra costs to companies. This phenomenon originates mainly from improper mud properties, bottomhole assembly (BHA) (contacting area), still pipe time, and differential pressure between the formation and the drilling mud. Investigation on various conditions that lead to DPS makes it possible to develop some preventive treatments to avoid this problem's occurrence. In the past, statistical methods were applied in this area, but recently artificial neural network (ANN) approaches are frequently being used. ANNs have some priorities over conventional statistical methods such as the model-free form of predictions, tolerance to data errors, data-driven nature, and fast computation. On the other hand, the designed ANNs have some shortcomings and restrictions as they are developed to predict problems. In this paper, to solve most of the existing disadvantages of ANNs, a novel support-vector machine (SVM) approach has been developed to predict a DPS occurrence in horizontal and sidetracked wells in Iranian offshore oil fields. The results from the analysis have shown the potential of the SVM and ANNs to predict DPS, with the SVM results being more promising.

Intelligent prediction of differential pipe sticking by support vector machine compared with conventional artificial neural networks: An example of iranian offshore oil fields
Reza Jahanbakhshi, Reza Keshavarzi, Mahdi Aliyari Shoorehdeli, Abolqasem Emamzadeh
SPE Drilling & Completion
2012Journal
Abstract:

This study presents fault tolerant control of inverted pendulum via on-line fuzzy backstepping and anti-control of chaos. The inverted pendulum, as a mechatronics system, is used frequently in robotic applications and can be found in different forms. Based on Lyapunov stability theory for backstepping design, the nonlinear controller and some generic sufficient conditions for asymptotic control are attained. Also in this study, anti-control of chaos is applied to increase the fault tolerant of inverted pendulum. To achieve this goal, the chaos dynamic must be created in the inverted pendulum system. So, the inverted pendulum system has been synchronized to chaotic gyroscope system. In this study, control and anti-control concepts are applied to achieve the high quality performance of inverted pendulum system. It is tried to design a controller which is capable to satisfy the control and anti- control aims. The performances of the proposed control are examined in terms of fault tolerant capability. Finally the efficacy of the proposed methods are illustrated by simulations.

Inverted Pendulum Fault Tolerant Control Based on Fuzzy Backstepping Design and Anti-Control of Chaos
Atefeh Saedian, Hassan Zarabadipour, Mahdi Aliyari Shoorehdeli, Faezeh Farivar
IFAC Proceedings Volumes
2012Journal
Abstract:

This paper presents a nonlinear control for trajectory tracking and vibration control of a flexible joint manipulator by using chaotic gyroscope synchronization. To study the effectiveness of the controllers, designed controller is developed for tip angular position control of a flexible joint manipulator. Based on Lyapunov stability theory, the nonlinear controller and some generic sufficient conditions for global asymptotic control are attained. In this study, the anti-control is applied to reduce the deflection angle of flexible joint system. To achieve this goal, the chaos dynamic must be created in the flexible joint system. So, the flexible joint system has been synchronized to chaotic gyroscope system. In this study, control and anti-control concepts are applied to achieve the high quality performance of flexible joint system. It is trying to design a controller which is capable to satisfy the control and anti-control aims. The performances of the proposed control are examined in terms of input tracking capability and level of vibration reduction. Finally, the efficacy of the proposed method is validated through experimentation on QUANSER’s flexible joint manipulator.

Lyapunov based control of flexible joint manipulator with experimental validation by using chaotic gyroscope synchronization
Mojtaba Rostami Kandroodi, Faezeh Farivar, Mahdi Aliyari Shoorehdeli
International Journal of Mechanic Systems Engineering
2012Journal
Abstract:

This study proposes a model-based robust fault detection and isolation (RFDI) method with hybrid structure. Robust detection and isolation of the realistic faults of an industrial gas turbine in steady-state conditions is mainly considered. For residual generation, a bank of time-delay multilayer perceptron (MLP) models is used, and in fault detection step, a passive approach based on model error modelling is employed to achieve threshold adaptation. To do so, local linear neuro-fuzzy (LLNF) modelling is utilised for constructing error-model to generate uncertainty interval upon the system output in order to make decision whether a fault occurred or not. This model is trained using local linear model tree (LOLIMOT) which is a progressive tree-construction algorithm. Simple thresholding is also used along with adaptive thresholding in fault detection phase for comparative purposes. Besides, another MLP neural network is utilised to isolate the faults. In order to show the effectiveness of proposed RFDI method, it was tested on a single-shaft industrial gas turbine prototype model and has been evaluated based on the gas turbine data. A brief comparative study with the related works done on this gas turbine benchmark is also provided to show the pros and cons of the presented RFDI method.

Model-based robust fault detection and isolation of an industrial gas turbine prototype using soft computing techniques
Hasan Abbasi Nozari, Mahdi Aliyari Shoorehdeli, Silvio Simani, Hamed Dehghan Banadaki
Neurocomputing
2012Journal
Abstract:

This paper proposes the modified projective synchronization for heavy symmetric dissipative gyroscope systems via backstepping control. Because of the nonlinear terms of the gyroscope system, the system exhibits complex and chaotic motions. Using the backstepping control technique, control laws are established which guarantees the hybrid projective synchronization including synchronization, anti-synchronization and projective synchronization. By Lyapunov stability theory, control laws are proposed to ensure the stability of the controlled closed-loop. Numerical simulations are presented to verify the proposed synchronization approach. This paper demonstrates that synchronization and anti- synchronization can coexist in dissipative gyroscope systems via nonlinear control.

Modified projective synchronization of chaotic dissipative gyroscope systems via backstepping control
Faezeh Farivar, Mohammad Ali Nekoui, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Indian Journal of Physics
2012Journal
Abstract:

This paper addresses the experimental identification of a servo actuator which is used in many industrial applications. Because the system consisted of electrical and mechanical components, the behavior of the system was nonlinear. In addition, the under load behavior of this servo was different. The load torque was considered as the input and a two input-one output model was presented for this servo actuator. Special focus was given in order to present a simple model for this servo actuator. The comparison between simulation and experimental results showed the effectiveness of the propose model. The model can be applied as a reference for characterizing different designs and future control strategies.

An under Load Servo Actuator Identification
Mohsen Maboodi, MH Ashtari Larki, M Aliyari Shoorehdeli, Hosein Bolandi
IFAC Proceedings Volumes
2011Journal
Abstract:

Overhead crane is an industrial structure that used widely in many harbors and factories. It is usually operated manually or by some conventional control methods. In this paper, we propose a hybrid controller includes both position regulation and anti-swing control. According to Takagi-Sugeno fuzzy model of an overhead crane and genetic algorithm, a fuzzy controller is designed with parallel distributed compensation and Linear Quadratic Regulation. Using genetic algorithm, important fuzzy rules are selected and so the number of rules decreased and design procedure need less computation and its computation needs less time. Further, the stability of the overhead crane with the parallel distributed fuzzy LQR controller is discussed. The stability analysis and control design problems is reduced to linear matrix inequality (LMI) problems. Simulation results illustrated the validity of the proposed parallel distributed fuzzy LQR control method and it was compared with a similar method parallel distributed fuzzy controller with same fuzzy rule set.

Anti-swing control for a double-pendulum-type overhead crane via parallel distributed fuzzy LQR controller combined with genetic fuzzy rule set selection
Mahdieh Adeli, Hassan Zarabadipour, Seyedeh Hamideh Zarabadi, Mahdi Aliyari Shoorehdeli
Control System, Computing and Engineering (ICCSCE), 2011 IEEE International Conference on
2011Conference
Abstract:

One of the common industrial structures that are used widely in many harbors and factories and buildings is overhead crane. Overhead cranes are usually operated manually or by some conventional control methods. In this paper, we propose a hybrid controller includes both position regulation and anti-swing control. According to Takagi-Sugeno fuzzy model of an overhead crane, a fuzzy controller designed with parallel distributed compensation and Linear Quadratic Regulation. With the Takagi-Sugeno fuzzy modeling, the nonlinear system is approximated by the combination of several linear subsystems in the corresponding fuzzy state space region. Then by constructing a linear quadratic regulation subcontroller according to each linear subsystem, a parallel distributed fuzzy LQR controller is designed. Further, the stability of the overhead crane with the parallel distributed fuzzy LQR controller is discussed. Simulation results illustrated the validity of the proposed control algorithm and it is compared with a similar method parallel distributed fuzzy controller.

Anti-swing control of a double-pendulum-type overhead crane using parallel distributed fuzzy LQR controller
Mahdieh Adeli, Hassan Zarabadipour, M Aliyari Shoorehdeli
Control, Instrumentation and Automation (ICCIA), 2011 2nd International Conference on
2011Conference
Abstract:

In the petroleum industry perforating is a method of making holes through the casing opposite the production formation to allow the oil or gas to flow into the well. In the current explosive shaped charge perforation method there arc some serious problems, such as producing debris. uncontrollable hole size and shape, compaction of rock formation in the area next to the tunnel and decreasing permeability. Recent advances in high power laser technology provide a new alternative to replace the current perforating gun. Due to the nature of oil and gas reservoirs, one of the challenges in laser perforation is the laser beam-fluid interaction that results in laser power loss (LPL), In this paper, feed-forward network with back-propagation and generalized regression neural networks have been developed to predict LPL in the laser beam-fluid interaction during laser perforation. Effective parameters in the laser-fluid interaction such as laser power, fluid viscosity and fluid thickness which arc related to laboratory tests done by ytterbium-doped multi-clad fibre laser are the inputs and LPL is the output of the neural networks. The developed neural networks have shown high correlation coefficients with low error and the LPL for the laser beam-fluid interaction during laser perforation was predicted with high accuracy.

Applying High Power Lasers in Perforating Oil and Gas Wells: Prediction of the Laser Power Loss During Laser Beam-Fluid Interaction by Using Artificial Neural Network
R Keshavarzi, R Jahanbakhshi, H Bayesteh, A Ghorbani, MA Shoorehdeli
Lasers in Engineering (Old City Publishing)
2011Journal
Abstract:

Overhead crane is an industrial structure that is widely used in many harbors and factories. It is usually operated manually or by some conventional control methods. In this paper, we propose a hybrid controller includes both position regulation and anti-swing control. According to Takagi-Sugeno fuzzy model of an overhead crane and genetic algorithm, a fuzzy controller is designed with parallel distributed compensation and Linear Quadratic Regulation. Using genetic algorithm, important fuzzy rules are selected and so the number of rules decreased and design procedure need less computation and its computation needs less time. Further, the stability of the overhead crane with the parallel distributed fuzzy LQR controller is discussed. The stability analysis and control design problems is reduced to linear matrix inequality (LMI) problems. Simulation results illustrated the validity of the proposed parallel distributed fuzzy LQR control method and it was compared with a similar method parallel distributed fuzzy controller with same fuzzy rule set.

Crane control via parallel distributed fuzzy LQR controller using genetic fuzzy rule selection
M Adeli, H Zarabadipour, M Aliyari Shoorehdeli
Control, Instrumentation and Automation (ICCIA), 2011 2nd International Conference on
2011Conference
Abstract:

Nowadays computer games have become a billion dollar industry. One of the important factors in success of a game is its similarity to the real world. As a result, many AI approaches have been exploited to make game characters more believable and natural. One of these approaches which has received great attention is Fuzzy Logic. In this paper a Fuzzy Rule-Based System is employed in a fighting game to reach higher levels of realism. Furthermore, behavior of two fighter bots, one based on the proposed Fuzzy logic and the other one based on a scripted AI, have been compared. It is observed that the results of the proposed method have less behavioral repetition than the scripted AI, which boosts human players' enjoyment during the game.

Deploying Fuzzy Logic in a Boxing Game
Hamid Reza Nasrinpour, Siavash Malektaji, M Aliyari Shoorehdeli, Mohammad Teshnehlab
Proceedings of the 6th Annual International North-American Conference on AI and Simulation in Games (GameON-NA), Troy, NY, USA
2011Conference
Abstract:

One of the common industrial structures that are used widely in many harbors and factories and buildings is overhead crane. Overhead cranes are usually operated manually or by some conventional control methods. In this paper, we propose a hybrid controller includes both position regulation and anti-swing control. According to Takagi-Sugeno fuzzy model of an overhead crane, a fuzzy controller designed with parallel distributed compensation and Linear Quadratic Regulation. With the Takagi-Sugeno fuzzy modeling, the nonlinear system is approximated by the combination of several linear subsystems in the corresponding fuzzy state space region. Then by constructing a linear quadratic regulation subcontroller according to each linear subsystem, a parallel distributed fuzzy LQR controller is designed. Further, the stability of the overhead crane with the parallel distributed fuzzy LQR controller is discussed. Simulation results illustrated the validity of the proposed control algorithm and it is compared with a similar method parallel distributed fuzzy controller.

Design of a parallel distributed fuzzy LQR controller for double-pendulum-type overhead cranes
Mahdieh Adeli, Seyedeh Hamideh Zarabadi, Hassan Zarabadipour, Mahdi Aliyari Shoorehdeli
Control System, Computing and Engineering (ICCSCE), 2011 IEEE International Conference on
2011Conference
Abstract:

This study presents a novel controller of magnetic levitation system by using new neuro-fuzzy structures which called flexible neuro-fuzzy systems. In this type of controller we use sliding mode control with neuro-fuzzy to eliminate the Jacobian of plant. At first, we control magnetic levitation system with Mamdanitype neuro-fuzzy systems and logical-type neuro-fuzzy systems separately and then we use two types of flexible neuro-fuzzy systems as controllers. Basic flexible OR-type neuro-fuzzy inference system and basic compromise AND-type neuro-fuzzy inference system are two new flexible neuro-fuzzy controllers which structure of fuzzy inference system (Mamdani or logical) is determined in the learning process. We can investigate with these two types of controllers which of the Mamdani or logical type systems has better performance for control of this plant. Finally we compare performance of these controllers with sliding mode controller and RBF sliding mode controller.

Designing flexible neuro-fuzzy system based on sliding mode Controller for magnetic levitation systems
Zahra Mohammadi, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
IJCSI International Journal of Computer Science Issues
2011Journal
Abstract:

Recently a lot of works have been done to detect faults in nonlinear systems. In this paper a new method, based on parity relations for linear systems, is proposed to detect faults in nonlinear systems that can be modeled by Takagi-Sugeno (TS) fuzzy system. This method is an intuitive generalization of parity relations, because TS fuzzy system uses local linear models. Results of simulation and implementation on a rotary inverted pendulum show that faults can be detected very well.

Fault detection of nonlinear systems by parity relations
Majid Ghaniee, Mahdi Aliyari Shoorehdeli
Control, Instrumentation and Automation (ICCIA), 2011 2nd International Conference on
2011Conference
Abstract:

In this paper, fault tolerant synchronization of chaotic gyroscope systems via Gaussian RBF neural network based on sliding mode control is investigated. Taking a general nature of fault in the slave system into account, a new synchronization scheme, namely, fault-tolerant synchronization, is proposed, by which the synchronization can be achieved no matter if the fault and disturbance occur or not. By making use of a slave-observer and Gaussian RBF Neural Network Based on Sliding Mode Control, the fault tolerant synchronization can be achieved. The adaptation law of designed controller is obtained based on sliding mode control methodology without calculating the Jacobian of the system. The proposed method can compensate the actuator faults and disturbances occurred in the slave system. Numerical simulation results demonstrate the validity and feasibility of the proposed method to fault tolerant synchronization.

Fault tolerant synchronization of chaotic heavy symmetric gyroscope systems via gaussian rbf neural network based on sliding mode control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli
Mechatronics (ICM), 2011 IEEE International Conference on
2011Conference
Abstract:

Breast cancer Dynamic magnetic resonance imaging (MRI) has emerged as a powerful diagnostic tool for breast cancer detection due to its high sensitivity and has established a role where findings from conventional mammography techniques are equivocal[1]. In the clinical setting, the ANN has been widely applied in breast cancer diagnosis using a subjective impression of different features based on defined criteria. In this study, feature selection and classification methods based on Artificial Neural Network (ANN) and Support Vector Machine (SVM) are applied to classify breast cancer on dynamic Magnetic Resonance Imaging (MRI). The database including benign and malignant lesions is specified to select the features and classify with proposed methods. It was collected from 2004 to 2006. A forward selection method is applied to find the best features for classification. Moreover, several neural networks classifiers like MLP, PNN, GRNN and RBF has been presented on a total of 112 histopathologically verified breast lesions to classify into benign and malignant groups. Also support vector machine have been considered as classifiers. Training and recalling classifiers are obtained with considering four-fold cross validation.

Feature selection and classification of breast cancer on dynamic magnetic resonance imaging using ANN and SVM
F Keyvanfard, MA Shoorehdeli, M Teshnehlab
American Journal of Biomedical Engineering
2011Journal
Abstract:

This paper aims to increase the classification specificity by using multi classifier system. First, a novel pixel search approach is applied to find significant region in images. Fuzzy C-means is utilized to determine the clear boundary of tumor. Then, shape and texture features are extracted from region of interest. Genetic algorithm is applied to select the best feature used for classifiers. Several neural networks and support vector machine are considered as classifiers that classify the data into benign and malignant group. To improve the performance of classification, three classifiers that have the best results among all applied methods are combined together that they have been named as multi classifier system. For each lesion, final detection as malignant or benign has been evaluated, when the same results are achieved from two classifiers of multi classifier system. Notice that the Jack-Knife technique is applied in this study, because it is useful for small data base as ours gotten from Milad Hospital in Tehran, Iran.

Feature selection and classification of breast MRI lesions based on Multi classifier
Farzaneh Keyvanfard, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Artificial Intelligence and Signal Processing (AISP), 2011 International Symposium on
2011Conference
Abstract:

This paper presents a new hybrid control strategy for velocity control of an electro hydraulic servo system (EHSS) in presence of flow nonlinearities and internal friction. We employed a combination of LQR controller and fuzzy-neural network in a feedback error learning framework. In the proposed control approach, LQR controller as a classical controller is designed such that the stability is guaranteed and the control purposes are satisfied. Then an intelligent controller (FNN) which is working with the classical controller (LQR) takes the control task completely. It is shown that this technique (fuzzy-LQR) has good performance and also it has a very fast and proper response. All derived results are validated by computer simulation of a nonlinear mathematical model of the system.

Fuzzy-LQR hybrid control of an electro hydraulic velocity servo system
Fereshteh Poloei, Maryam Zekri, Mahdi Aliyari Shoorehdeli
Hybrid Intelligent Systems (HIS), 2011 11th International Conference on
2011Conference
Abstract:

This paper proposes the generalized projective synchronization (GPS) of uncertain chaotic systems with external disturbance via Gaussian radial basis adaptive sliding mode control (GRBASMC). A sliding surface is adopted to ensure the stability of the error dynamics in sliding mode control. In the neural sliding mode controller, a Gaussian radial basis function is utilized to online estimate the system dynamic function. The adaptation law of the control system is derived in the sense of Lyapunov function, thus the system can be guaranteed to be asymptotically stable. The proposed method allows us to arbitrarily adjust the desired scaling by controlling the slave system. It is not necessary to calculate the Lyapunov exponents and the eigen values of the Jacobian matrix, which makes it simple and convenient. Also, it is a systematic procedure for GPS of chaotic systems and it can be applied to a variety of chaotic systems no matter whether it contains external excitation or not. Note that it needs only one controller to realize GPS no matter how much dimensions the chaotic system contains and the controller is easy to be implemented. The proposed method is applied to three chaotic systems: Genesio system, Lur’e like system and Duffing system.

Generalized projective synchronization of uncertain chaotic systems with external disturbance
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
Expert Systems with Applications
2011Journal
Abstract:

In this study, Extended Kalman Filter (EKF) algorithm is developed to estimate the parameters of Hammerstein-Wiener (HW) ARMAX models. The basic idea is to estimate the original parameters of the identification model, which are appeared in the form of product terms, directly. While, other algorithms like Extended Forgetting Factor Stochastic Gradient (EFG), Extended Stochastic Gradient (ESG), Forgetting Factor Recursive Least Square (FFRLS) and Kalman Filter (KF), estimate parameters in the product form and they need another algorithms such averaging method (AVE method), singular value decomposition method (SVD method) to separate the parameters. So, the computational complexity of the proposed approach decreases. To show the efficiency of this method the results are compared with EFG and ESG method.

Identification of hammerstein-wiener armax systems using extended kalman filter
M Mansouri, H Tolouei, M Aliyari Shoorehdeli
Control and Decision Conference (CCDC), 2011 Chinese
2011Conference
Abstract:

In this paper, a new algorithm is presented in using Multi Layer Perceptron (MLP) and Radial Base Function (RBF) to predict Ischemia diseases by Electrocardiogram (ECG) signals. The process would be very difficult due to non-stationary and nonlinear characteristics of ECG signals. MLP and RBF algorithms are well known in predicting the problems. However, they have not been used for real time prediction through signals, especially bio signals such as ECG. Pre-processing is necessary for ECG signal in order to detect QRS complex. Regarding the extract influential features in Ischemia disease, the baseline wandering and noise suppression are done. MLP and RBF, the predictors, are employed to foresee the further next beats in ECG signals. The validity of predictor accuracy is evaluated by Root Mean Square Error (RMSE) criterion. After the prediction stage, The predicted beats are classified by Adaptive Neuro-Fuzzy network (ANFIS) classifier as ischemic and normal. MLP and RBF are tested for their abilities in order to predict Ischemic Heart Disease (IHD) upon ECG signals. The performances of classified beats are evaluated based on computed Sensitivity (Se) and Specificity (Sp). In this study several ECG signals recorded by European Society of Cardiology for ST-T database are used. By applying prediction methods (Direct and Recursive Predictions) 48 steps can be predicted ahead in ECG signal. Then the predicted beats are classified as Ischemic or normal beats. Therefore, the ischemic beats can be predicted in 48 steps ahead. By comparing the results obtained in this study, the MLP and RBF networks are evaluated for their capabilities in predicting Ischemia. According to this comparison, MLP shows better results and the results of ANFIS as a classifier has been satisfactory enough in classification of Ischemic beats. Therefore, these results can be used for early diagnosis of Ischemic Heart Disease (IHD).

Ischemia prediction via ECG using MLP and RBF predictors with ANFIS classifiers
Hoda Tonekabonipour, Ali Emam, Mohamad Teshnelab, Mahdi Aliyari Shoorehdeli
Natural Computation (ICNC), 2011 Seventh International Conference on
2011Conference
Abstract:

This study proposed a model based fault detection and isolation (FDI) method using multi- layer perceptron (MLP) neural network. Detection and isolation of realistic faults of an industrial gas turbine engine in steady-state conditions is mainly centered. A bank of MLP models which are obtained by nonlinear dynamic system identification is used to generate the residuals, and also simple thresholding is used for the intend of fault detection while another MLP neural network is employed to isolate the faults. The proposed FDI method was tested on a single-shaft industrial gas turbine prototype and it have been evaluated using non-linear simulations based on the real gas turbine data. A brief comparative study with other related works in the literature on this gas turbine benchmark is also provided to show the benefits of proposed FDI method.

Model-based fault detection and isolation using neural networks: An industrial gas turbine case study
Hasan Abbasi Nozari, Hamed Dehghan Banadaki, Mehdi Aliyari Shoorehdeli, Silvio Simani
Systems Engineering (ICSEng), 2011 21st International Conference on
2011Conference
Abstract:

This paper presents a Neuro-fuzzy based method using local linear model trees (LOLIMOT) train algorithm for nonlinear identification of a catalytic reformer unit in oil refinery plant. This unit include highly nonlinear behaviour and it is complicated to obtain an accurate physical model. There for, it is necessary to use such appropriate method providing suitable while preventing computational complexities. LOLIMOT algorithm as an incremental learning algorithm has been used several time as a well-known method for nonlinear system identification and estimation. For comparison, Multi Layer Perceptron (MLP) and Radial Bases Function (RBF) neural networks as well-known methods for nonlinear system identification and estimation are used to evaluate the performance of LOLIMOT. The results presented in this paper clearly demonstrate that the LOLIMOT is superior to other methods in identification of nonlinear system such as catalytic reformer unit (CRU).

Modeling and identification of catalytic reformer unit using locally linear model trees
Mohammad Mokhtare, Somayeh Hekmati Vahed, Mahdi Aliyari Shoorehdeli, Alireza Fatehi
Electrical Engineering (ICEE), 2011 19th Iranian Conference on
2011Conference
Abstract:

The aim of this study is to prove validity of feedback error learning rule for a linear representation of dynamic system with unknown parameters. A simple single-layer neural network is assumed as an adaptive linear combiner and stability techniques are applied to derive the same adaptation law as feedback error learning rule.

Stability of feedback error learning for linear systems
Mehdi Tavan, Mahdi Aliyari Shoorehdeli, Amir Reza Zare Bidaki
IFAC Proceedings Volumes
2011Journal
Abstract:

This study addresses a nonlinear trajectory tracking control problem for a kinematics Model of nonholonomic mobile robot with considering next 2 time path curvature. The tracking control of mobile robot using two cascade controllers is presented. The first fuzzy controller produces a variable which shows curvature of the path and is considered as one of the inputs of the second fuzzy controller. Adaptive Neuro Fuzzy Inference System (ANFIS) is applied as second stage controller for the solution of the path tracking problem of mobile robots. The inputs value to fuzzy logic layer are VC, C, dR & dθ the robot current linear velocity, trajectory curvature, distance from the robot actual position to the next desired position, and difference between the angles of the dθ and the robot actual heading, respectively. A gradient descent learning algorithm is used to adjust the parameters. That present controller is compared with previous work to confirm its effectiveness.

Tracking control of mobile robot using ANFIS
Masoud Imen, Mohammad Mansouri, Mehdy Aliyari Shoorehdeli
Mechatronics and Automation (ICMA), 2011 International Conference on
2011Conference
Abstract:

This paper presents a variable structure control and anti control for trajectory tracking and vibration control of a flexible joint manipulator. To study the effectiveness of the controllers, designed controller is developed for tip angular position control of a flexible joint manipulator. Based on Lyapunov stability theory for variable structure control, the nonlinear controller and some generic sufficient conditions for global asymptotic control are attained. Also in this study, the anti-control is applied to reduce the deflection angle of flexible joint system. To achieve this goal, the chaos dynamic must be created in the flexible joint system. So, the flexible joint system has been synchronized to chaotic gyroscope system. In this study, control and anti-control concepts are applied to achieve the high quality performance of flexible joint system. It is tried to design a controller which is capable to satisfy the control and anti-control aims. The performances of the proposed control are examined in terms of input tracking capability, level of vibration reduction, and time response specifications. Finally, the efficacy of the proposed method is validated through experimentation on QUANSER's flexible-joint manipulator.

Variable structure control and anti-control of flexible joint manipulator with experimental validation
Mojtaba Rostami Kandroodi, Faezeh Farivar, Maysam Zamani Pedram, Mahdi Aliyari Shoorehdeli
Mechatronics (ICM), 2011 IEEE International Conference on
2011Conference
Abstract:

The use of intravenous drugs in general anesthesia is increasingly popular. Because of relatively precise injection rate, the goal of consistent control is possible. Because of using different drugs in full anesthesia for adequate hypnosis, analgesia and muscle relaxation, the interaction between drugs is more considerable especially when intravenous drugs are used. In this paper we use a developed Pharmacokinetic-Pharmacodynamic model which considers the interaction between two more popular intravenous drugs, Propofol for hypnosis and Remifentanil as an analgesic drug, to design a closed-loop system. The Radial Basis Function (RBF) controller as an adaptive neural controller was designed and adaptive properties of this structure in confront of variations in model parameters values was investigated. Trying to improve the tracking performance, one of most popular methods in hybrid control, Feedback Error Learning (FEL), was utilized.

A new approach in drug delivery control in anesthesia
M Aliyari, M Teshnehlab
Systems Man and Cybernetics (SMC), 2010 IEEE International Conference on
2010Conference
Abstract:

In this study, adaptive control of flexible link model which is non-minimum phase and single- input, multiple-output (SIMO) is presented. The controllers designed aim to control the hub position in a way that attenuates the tip deflections with less energy consumption. Methods used to design the under actuated controller are WRBF network and neuro-fuzzy network and are compared to LQR and non-adaptive fuzzy controller. Learning method performed for adaptive schemes is emotional. Simulation results show the effectiveness of the designed controllers and reduction of energy demand in intelligent adaptive controllers.

Adaptive intelligent control of flexible link robot arm
Nassim Nikpay, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Intelligent Systems and Informatics (SISY), 2010 8th International Symposium on
2010Conference
Abstract:

This paper presents a new approach for breast cancer detection based on Hierarchical Fuzzy Neural Network (HFNN). Generally in formal fuzzy neural networks (FNN), increasing the number of inputs, causes exponential growth in the number of parameters of the FNN system. This phenomenon named as" curse of dimensionality". An approach to deal with this problem is to use the hierarchical fuzzy neural network. A HFNN consists of hierarchically connected low-dimensional fuzzy neural networks. HFNN can use less rules to model nonlinear system. This method is applied to the Wisconsin Breast Cancer Database (WBCD) to classify breast cancer into two groups: benign and malignant lesions. The performance of HFNN is then compared with FNN by using the same breast cancer dataset.

Breast cancer detection by using hierarchical fuzzy neural system with EKF trainer
Seyedeh Somayeh Naghibi, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
2010 17th Iranian Conference of Biomedical Engineering (ICBME)
2010Conference
Abstract:

Fuzzy modeling of high-dimensional systems is a challenging topic. This study proposes an effective approach to data-based fuzzy modeling of high-dimensional systems. The proposed method works on the fuzzification layer and tries to use two-dimensional membership functions instead of onedimensional ones. This approach reduces fuzzy rule base radically due to using of two-dimensional membership functions which lead to reduction of parameters. The resulting fuzzy system generated by this method has the following distinct features: 1) the fuzzy system is quite simplified; 2) the fuzzy system is interpretable; 3) the dependencies between the inputs and the outputs are clearly shown. This method has successfully been applied to three classification problem and the results are compared with other works.

Classification of Multi-Class Datasets Using 2D Membership Functions in TSK Fuzzy System.
Loghman Kaki, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
Int. J. Adv. Comp. Techn.
2010Journal
Abstract:

This paper describes hybrid multivariate method: Principal Component Analysis improved by Genetic Algorithm. This method determines main Principal Components can be used to detect fault during the operation of industrial process by neural classifier. This technique is applied to simulated data collected from the Tennessee Eastman chemical plant simulator which was designed to simulate a wide variety of faults occurring in a chemical plant based on a facility at Eastman chemical.

Fault detection of the Tennessee Eastman process using improved PCA and neural classifier
Mostafa Noruzi Nashalji, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Soft computing in industrial applications
2010Journal
Abstract:

In this paper, a new feature selection and classification methods based on artificial neural network are applied to classify breast cancer on dynamic Magnetic Resonance Imaging (MRI). The database including benign and malignant lesions is specified to select the features and classify with proposed methods. It is collected from 2004 to 2006.

Feature selection and classification of breast cancer on dynamic Magnetic Resonance Imaging by using artificial neural networks
Farzaneh Keivanfard, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli, Ke Nie, Min-Ying Su
Biomedical Engineering (ICBME), 2010 17th Iranian Conference of
2010Conference
Abstract:

This paper proposes a neural sliding mode control scheme for the synchronization of two chaotic nonlinear gyros subject to uncertainties and external disturbance. In this scheme, sliding mode control and multi layer perceptron neural network control are combined. A sliding surface is adopted to ensure the stability of the error dynamics in sliding mode control. The adaptation law of the multi layer perceptron neural network control system is derived in the sense of Lyapunov function, thus the system can be guaranteed to be asymptotically stable. By Lyapunov stability theory, neural sliding mode control is presented to ensure the stability of the controlled system. Multi layer perceptron Neural Network control is trained during the control process. The proposed method allows us to synchronize gyros by controlling the slave system. It is not necessary to calculate the Lyapunov exponents and the eigenvalues of the jacobian matrix, which makes it simple and convenient. Also, it is a systematic procedure for chaos synchronization of uncertain nonlinear gyro systems. Note that it needs only one controller to realize synchronization and the controller is easy to be implemented. The simulation results demonstrate the ability of the neural sliding control scheme to synchronize the chaotic gyro systems.

Neural Sliding Mode Control for Chaos Synchronization of Uncertain Nonlinear Gyros
Faezeh Farivar, Mohammad Ali Nekoui, Mahdi Aliyari Shoorehdeli
Advances and Applications in Mathematical Sciences
2010Journal
Abstract:

A number of techniques for detection of faults in ball bearing using frequency domain approach exist today. For analyzing non-stationary signals arising out of defective rolling element bearings, use of conventional discrete Fourier Transform (DFT) has been known to be less efficient. One of the most suited time-frequency approach; wavelet transform (WT) has inherent problems of large computational time and fixed-scale frequency resolution. In view of such constraints, the Hilbert-Huang Transform (HHT), technique provides multi-resolution in various frequency scales and takes the signal's frequency content and their variation into consideration. HHT analyses the vibration signal using intrinsic mode functions (IMFs), which are extracted using the process of empirical mode decomposition (EMD). HHT is effective in many different fields but lacks proper theoretical support. The time resolution significantly affects the calculation of corresponding frequency content of the signal. In this paper Firstly, the EMD method is used as a pretreatment to decompose the non-stationary vibration signal of a roller bearing into a number of intrinsic mode function (IMF) components which are stationary. Secondly, we choose some special IMFs to obtain Hilbert transform and then Hilbert marginal spectrum and the last local Hilbert marginal spectrum from which the faults in a roller bearing can be diagnosed and fault patterns can be identified. Finally, the characteristic amplitude ratios serve as the fault characteristic vectors to be input to the multi-class support vector machine (MSVM) classifiers and the work condition and fault patterns of the roller bearings and then faults are diagnosis real time based on Voting.

Notice of Retraction Multi-fault diagnosis of ball bearing using intrinsic mode functions, Hilbert marginal spectrum and multi-class support vector machine
OR Seryasat, M Aliyari Shoorehdeli, F Honarvar, A Rahmani, J Haddadnia
Mechanical and Electronics Engineering (ICMEE), 2010 2nd International Conference on
2010Conference
Abstract:

One of the most common cardiovascular diseases is Myocardial Ischemia (MI). The aim of this study is improving the diagnosis level of Ischemic Beat detection from ECG signals which is important in ischemic episode detection process. This improvement is based on appropriate feature extraction via Multi resolution Wavelet analysis and proper classifier selection. The approach starts with signal preprocessing, Afterwards efficacious morphologic features which are useful in ischemia detection are extracted by wavelet analysis. In the third stage subtractive clustering is performed for clustering. Finally probabilistic neural networks are used as a classifier. The proposed algorithm is evaluated on the European Society of Cardiology (ESC) ST-T database and reported 96.67% sensitivity and 89.18% specificity.

Probabilistic neural network oriented classification methodology for ischemic beat detection using multi resolution wavelet analysis
Shiva Khoshnoud, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
Biomedical Engineering (ICBME), 2010 17th Iranian Conference of
2010Conference
Abstract:

It is widely accepted in the brain computer interface research community that neurological phenomena are the only source of control in any BCI system. Artifacts are undesirable signals that can interfere with neurological phenomena. They may change the characteristics of neurological phenomena or even be mistakenly used as the source of control in BCI systems. Independent component analysis is a method that blindly separates mixtures of independent source signals, forcing the components to be independent. It has been widely applied to remove artifacts from EEG signals. Preliminary studies have shown that ICA increases the strength of motor-related signal components in the Mu rhythms, and is thus useful for removing artifacts in BCI systems.

An evolutionary artifact rejection method for brain computer interface using ICA
A Asadi Ghanbari, MR Nazari Kousarrizi, M Teshnehlab, M Aliyari
International Journal of Electrical & Computer Sciences
2009Journal
Abstract:

Monitoring and controlling the depth of anesthesia in surgery is so important. Compartmental models are well suited for closed-loop control of drug administration. In this paper, we develop a neural network and a fuzzy controller for nonlinear and compartmental system with nonnegative control input. In addition, the controllers guarantee that the physical system states remain in the nonnegative state space. After that, the proposed approaches are used to control the infusion of the anesthetic drug propofol in order to maintain a desired constant level of the depth of anesthesia for noncardiac surgery. In the end, this goal can be reached that intelligent systems are better than classic adaptive controller in adjustment of anesthesia with suitable condition of patient.

Anesthesia control based on intelligent controllers
N Eshghi, M Aliyari, M Teshnehlab
Bioinformatics and Biomedical Engineering, 2009. ICBBE 2009. 3rd International Conference on
2009Conference
Abstract:

This paper proposes a hybrid control scheme for the synchronization of two chaotic nonlinear gyros, subject to uncertainties and external disturbances. In this scheme, Linear Quadratic Regulation (LQR) control, Sliding Mode (SM) control and Gaussian Radial basis Function Neural Network (GRBFNN) control are combined. By Lyapunov stability theory, SM control is presented to ensure the stability of the controlled system. GRBFNN control is trained during the control process. The learning algorithm of the GRBFNN is based on the minimization of a cost function which considers the sliding surface and control effort. Simulation results demonstrate the ability of the hybrid control scheme to synchronize the chaotic gyro systems through the application of a single control signal.

Chaos synchronization of uncertain nonlinear gyros via hybrid control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
Advanced Intelligent Mechatronics, 2009. AIM 2009. IEEE/ASME International Conference on
2009Conference
Abstract:

The alcoholism is one of psychiatric phenotype, which results from interplay between genetic and environmental factors. Not only it leads to brain defects but also associated cognitive, emotional, and behavioral impairments. It can be detected by analyzing EEG signals. In this research, the power spectrum of the Haar mother wavelet is extracted as features. Then the principle component analysis is applied for dimension reduction of the feature vectors. Finally support vectors machine and neural networks are used for classification. The simulation results show that our proposed method achieves better classification accuracy than the other methods.

Classification of alcoholics and non-alcoholics via EEG using SVM and neural networks
Mohammad Reza Nazari Kousarrizi, Abdolreza Asadi Ghanbari, Ali Gharaviri, Mohammad Teshnehlab, Mahdi Aliyari
Bioinformatics and Biomedical Engineering, 2009. ICBBE 2009. 3rd International Conference on
2009Conference
Abstract:

Past work on face detection has emphasized the issues of feature extraction and classification, however, less attention has been given on the critical issue of feature selection. We consider the problem of face and non-face classification from frontal facial images using feature selection and neural networks. We argue that feature selection is an important issue in face and non-face classification. Automatic feature subset selection distinguishes the proposed method from previous face classification approaches. First, Principal Component Analysis (PCA) is used to represent each image as a feature vector (i.e., eigen-features) in a low-dimensional space, spanned by the eigenvectors of the covariance matrix of the training images (i.e., coefficients of the linear expansion).Then we consider Linear Discrimination Analysis (LDA) to achieve a comparison result between these two methods of dimension reduction. Genetic Algorithm (GA) is then used to select a subset of features from the low-dimensional representation by removing certain eigenvectors that do not seem to encode important information about face. Finally, a Probabilistic Neural Network (PNN) is trained to perform face classification using the selected eigen-feature subset. Experimental results demonstrate a significant improvement in error rate reduction.

Face detection based on dimension reduction using probabilistic neural network and genetic algorithm
Afsaneh Alavi Naini, Fatemeh Seiti, Mohammad Teshnelab, Mahdi Aliyari Shoorehdeli
Mechatronics and its Applications, 2009. ISMA'09. 6th International Symposium on
2009Conference
Abstract:

Brain Computer Interface one of hopeful interface technologies between humans and machines. Electroencephalogram-based Brain Computer Interfaces have become a hot spot in the research of neural engineering, rehabilitation, and brain science. The artifacts are disturbance that can occur during the signal acquisition and that can alter the analysis of the signals themselves. Detecting artifacts produced in electroencephalography data by muscle activity, eye blinks and electrical noise is a common and important problem in electroencephalography research. In this research, we used five different methods for detecting trials containing artifacts. Finally we used two different neural networks, and support vector machine to classify features that are extracted by wavelet transform.

Feature extraction and classification of EEG signals using wavelet transform, SVM and artificial neural networks for brain computer interfaces
Mohammad Reza Nazari Kousarrizi, AbdolReza Asadi Ghanbari, Mohammad Teshnehlab, Mahdi Aliyari Shorehdeli, Ali Gharaviri
Bioinformatics, Systems Biology and Intelligent Computing, 2009. IJCBS'09. International Joint Conference on
2009Conference
Abstract:

This study proposes a Gaussian Radial Basis Adaptive Backstepping Control (GRBABC) system for a class of n-order nonlinear systems. In the neural backstepping controller, a Gaussian radial basis function is utilized to on-line estimate of the system dynamic function. The adaptation laws of the control system are derived in the sense of Lyapunov function, thus the system can be guaranteed to be asymptotically stable. The proposed GRBABC is applied to two nonlinear chaotic systems which have the different order to illustrate its effectiveness. Simulation results verify that the proposed GRBABC can achieve favorable tracking performance by incorporating of GRBFNN identification, adaptive backstepping control techniques.

Gaussian radial basis adaptive backstepping control for a class of nonlinear systems
Faezeh Farivar, M Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
Applied Sci
2009Journal
Abstract:

This paper proposes the generalized projective synchronization for chaotic systems via Gaussian Radial Basis Adaptive Backstepping Control. In the neural backstepping controller, a Gaussian radial basis function is utilized to on-line estimate the system dynamic function. The adaptation laws of the control system are derived in the sense of Lyapunov function, thus the system can be guaranteed to be asymptotically stable. The proposed method allows us to arbitrarily adjust the desired scaling by controlling the slave system. It is not necessary to calculate the Lyapunov exponents and the eigen values of the Jacobian matrix, which makes it simple and convenient. Also, it is a systematic procedure for generalized projective synchronization of chaotic systems and it can be applied to a variety of chaotic systems no matter whether it contains external excitation or not. Note that it needs only one controller to realize generalized projective synchronization no matter how much dimensions the chaotic system contains and the controller is easy to be implemented. The proposed method is applied to three chaotic systems: Genesio system, Rössler system, and Duffing system.

Generalized projective synchronization for chaotic systems via Gaussian radial basis adaptive backstepping control
Faezeh Farivar, Mahdi Aliyari Shoorehdeli, Mohammad Ali Nekoui, Mohammad Teshnehlab
Chaos, Solitons & Fractals
2009Journal
Abstract:

This study describes hybrid control methods to control a flexible manipulator with payload. The dynamic equation of the system has been derived by Lagrange`s method. The designed controllers consist of two parts, classical controllers, PID and Linear Quadratic Regulation (LQR) and hybrid controllers, Fuzzy Neural Network (FNN) controller with Feedback Error Learning (FEL) and Sliding mode control using Gaussian Radial Basis Function Neural Network (RBFNN). The fuzzy neural network and radial basis function neural network are trained during control process and they are not necessarily trained off-line.

Hybrid control of flexible manipulator
F Farivar, M Aliyari Shoorehdeli, M Teshnehlab, MA Nekoui
Journal of Applied Sciences
2009Journal
Abstract:

This paper suggests novel hybrid learning algorithm with stable learning laws for adaptive network based fuzzy inference system (ANFIS) as a system identifier and studies the stability of this algorithm. The new hybrid learning algorithm is based on particle swarm optimization (PSO) for training the antecedent part and gradient descent (GD) for training the conclusion part. Lyapunov stability theory is used to study the stability of the proposed algorithm. This paper, studies the stability of PSO as an optimizer in training the identifier, for the first time. Stable learning algorithms for the antecedent and consequent parts of fuzzy rules are proposed. Some constraints are obtained and simulation results are given to validate the results. It is shown that instability will not occur for the leaning rate and PSO factors in the presence of constraints. The learning rate can be calculated on-line and will provide an adaptive learning rate for the ANFIS structure. This new learning scheme employs adaptive learning rate that is determined by input–output data.

Identification using ANFIS with intelligent hybrid stable learning algorithm approaches
Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab, Ali Khaki Sedigh
Neural Computing and Applications
2009Journal
Abstract:

This paper proposes a novel hybrid learning algorithm with stable learning laws for Adaptive Network based Fuzzy Inference System (ANFIS) as a system identifier and studies the stability of this algorithm. The new hybrid learning algorithm is based on particle swarm optimization (PSO) for training the antecedent part and forgetting factor recursive least square (FFRLS) for training the conclusion part. Two famous training algorithms for ANFIS are the gradient descent (GD) to update antecedent part parameters and using GD or recursive least square (RLS) to update conclusion part parameters. Lyapunov stability theory is used to study the stability of the proposed algorithms. This paper, also studies the stability of PSO as an optimizer in training the identifier. Stable learning algorithms for the antecedent and consequent parts of fuzzy rules are proposed. Some constraints are obtained and simulation results are given to validate the results. It is shown that instability will not occur for the leaning rate and PSO factors in the presence of constraints. The learning rate can be calculated on-line and will provide an adaptive learning rate for the ANFIS structure. This new learning scheme employs adaptive learning rate that is determined by input–output data. Also, stable learning algorithms for two common methods are proposed based on Lyapunov stability theory and some constraints are obtained.

Identification using ANFIS with intelligent hybrid stable learning algorithm approaches and stability analysis of training methods
Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab, Ali Khaki Sedigh, M Ahmadieh Khanesar
Applied Soft Computing
2009Journal
Abstract:

This paper suggests a novel approach for control of a flexible-link based on the feedback- error-learning (FEL) strategy. A radial basis function neural network (RBFNN) is used as an adaptive controller to improve the performance of a lead compensator controller in FEL structure. This scheme is developed by using a modified version of the FEL approach to learn the inverse dynamic of the flexible manipulator which requires only a linear model of the system for designing lead compensators and RBFNN controllers. The final controller should allow the user to command a desired tip angle position. The controller should eliminate the link's vibrations while maintaining a desirable level of response. Finally, the control performance of the proposed control approach for tip position tracking of flexible-link manipulator is illustrated by simulation result.

New control strategy of feedback error learning based on lead compensator for flexible link manipulator
Veser Namazikhah, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Control Conference (ECC), 2009 European
2009Conference
Abstract:

Thyroid gland produces thyroid hormones to help the regulation of the body's metabolism. The abnormalities of producing thyroid hormones are divided into two categories. Hypothyroidism which is related to production of insufficient thyroid hormone and hyperthyroidism related to production of excessive thyroid hormone. Separating these two diseases is very important for thyroid diagnosis. Therefore support vector machines and probabilistic neural network are proposed to classification. These methods rely mostly on powerful classification algorithms to deal with redundant and irrelevant features. In this paper feature selection is argued as an important problem via diagnosis and demonstrate that GAs provide a simple, general and powerful framework for selecting good subsets of features leading to improved diagnosis rates. Thyroid disease datasets are taken from UCI machine learning dataset.

Thyroid disease diagnosis based on genetic algorithms using PNN and SVM
Fatemeh Saiti, Afsaneh Alavi Naini, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Bioinformatics and Biomedical Engineering, 2009. ICBBE 2009. 3rd International Conference on
2009Conference
Abstract:

This paper proposes a novel hybrid learning algorithm with stable learning laws for Adaptive Network-based Fuzzy Inference System (ANFIS) as a system identifier. The proposed hybrid learning algorithm is based on the particle swarm optimization (PSO) for training the antecedent part and the extended Kalman filter (EKF) for training the conclusion part. Lyapunov stability theory is used to study the stability of the proposed algorithm. Comparison results of the proposed approach, PSO algorithm for training the antecedent part and recursive least squares (RLSs) or EKF algorithm for training the conclusion part, with the other classical approaches such as, gradient descent, resilient propagation, quick propagation, Levenberg–Marquardt for training the antecedent part and RLSs algorithm for training the conclusion part are provided. Moreover, it is shown that applying PSO, a powerful optimizer, to optimally train the parameters of the membership function on the antecedent part of the fuzzy rules in ANFIS system is a stable approach which results in an identifier with the best trained model. Stability constraints are obtained and different simulation results are given to validate the results. Also, the stability of Levenberg–Marquardt algorithms for ANFIS training is analyzed.

Training ANFIS as an identifier with intelligent hybrid stable learning algorithm based on particle swarm optimization and extended Kalman filter
Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab, Ali Khaki Sedigh
Fuzzy Sets and Systems
2009Journal
Abstract:

In this paper a Mamdani type fuzzy system and an adaptive network based fuzzy inference system (ANFIS) are presented for velocity control of an electro hydraulic servo system (EHSS) in presence of flow nonlinearities and internal friction. The architecture and learning procedure ANFIS is presented, which is a fuzzy inference system implemented in the framework of adaptive networks. It is shown that both these controllers can be successfully used to stabilize any chosen operating point of the system. All derived results are validated by computer simulation of nonlinear mathematical model of the system.

Velocity control of EHSS by using Mamdani and ANFIS controllers
Veeda Aghaei Hesari, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Control Conference (ECC), 2009 European
2009Conference
Abstract:

Brain Computer Interface (BCI) is a technology that developed over the last three decades has provided a novel and promising alternative method for interacting with the environment. BCI is a system which translates a subject's intentions into a control signal for a device, e.g., a computer application, a wheelchair or a neuroprosthesis. Electroencephalogram-based BCI has become a hot spot in the research of neural engineering, rehabilitation, and brain science. The artifacts are disturbance that can occur during the signal acquisition and that can alter the analysis of the signals themselves. Removing artifacts produced in Electroencephalogram (EEG) data by muscle activity, eye blinks and electrical noise is a common and important problem in EEG analysis. In this research, for artifact rejection, EEG data are filtered to the frequency range between 8 and 32 Hz with a butterworth band-pass filter. Finally two different structures of neural network and a support vector machine used to classify features that are extracted by Hilbert and Wavelet transform.

Wavelet and Hilbert transform-based brain computer interface
A Asadi Ghanbari, MR Nazari Kousarrizi, M Teshnehlab, M Aliyari
Advances in Computational Tools for Engineering Applications, 2009. ACTEA'09. International Conference on
2009Conference
Abstract:

Abstract A method of using particle swarm optimization (PSO) algorithm to design electromagnetic absorber is presented. To demonstrate effectiveness of the PSO algorithm three different design cases are optimized. To reduce the local minimum traps, a modified local search strategy is employed. Each design problem is optimized using genetic algorithm (GA) and four variants of PSO algorithms, namely global PSO (gbest), local PSO (lbest), comprehensive learning PSO (CLPSO), and modified local PSO (MLPSO). The results clearly show that the MLPSO is a robust, fast, and useful optimization tool for designing absorbers. A seven-layer absorber achieved by this method has reflection coefficient below 18.7 dB from VHF to 20 GHz.

Design of very thin wide band absorbers using modified local best particle swarm optimization
Somayyeh Chamaani, Seyed Abdullah Mirtaheri, Mahdi Aliyari Shooredeli
AEU-International Journal of Electronics and Communications
2008Journal
Abstract:

In the present research, a non-linear controller is designed for the control of an active suspension system for a half-model vehicle, using a Fuzzy Neural Network (FNN) along with Feedback error learning. The purpose in a vehicle suspension system is reduction of transmittance of vibrational effects from the road to the vehicle chassis, hence providing ride comfort. This requires a minimum reduction in road contact along rough roads. In addition, the role of the suspension system in vehicle control along a curved route and in accelerating and braking is quite evident. To accomplish this, one can first design a PD controller for the suspension system, using a classic control method and use it to train a fuzzy controller. This controller can be trained using the PD controller output error on an online manner. Once trained, the PD controller is removed from the control loop and the neuro-fuzzy controller takes on. In case of a change in the parameters of the system under control, the PD controller enters the control loop again and the neural network gets trained again for the new condition. Important characteristics of the proposed controller is that no mathematical model is needed for the system components, such as the non-linear actuator, spring, or shock absorber, and that no system Jacobian is needed. The performance of the proposed FNN controller is compared with that of the PD controller through simulations. The results show that the proposed controller is indeed capable of meeting the stated control requirements.

Designing a neuro-fuzzy controller for a vehicle suspension system using feedback error learning
SH Sadati, M Aliyari Shooredeli, AD Panah
Journal of mechanics and aerospace
2008Journal
Abstract:

In This paper, a new adaptive controller based on modified feedback error learning (FEL) approaches is proposed for load frequency control (LFC) problem. The FEL controller consists of neural network feedforward controller (NNFC) and conventional feedback controller (CFC), where the CFC is essential to guarantee global asymptotic stability of the overall system. Also, for improved the performance of system the dynamic neural network (DNN) is adopted in NNFC instead of conventional neural network. This neural network has dynamic in its structure and consists of two units: inhibitory and excitatory unit. The proposed FEL controller has been compared with the conventional FEL (CFEL) controller and the PID controller through some performance indices.

Dynamic Neural Network by using Feedback Error Learning Approaches for LFC in Interconnected Power System
K Sabahi, M Tehnehlab, M Aliyari, K Mahdizadeh
Iranian Conference on Electrical Engineering, 2008
2008Conference
Abstract:

In this study, a new group method of data handling (GMDH) method, based on adaptive neurofuzzy inference system (ANFIS) structure, called ANFIS-GMDH and its application for diabetes mellitus forecasting is presented. Conventional neurofuzzy GMDH (NF-GMDH) uses radial basis network (RBF) as the partial descriptions. In this study the RBF partial descriptions are replaced with two input ANFIS structures and backpropagation algorithm is chosen for learning this network structure. The Prima Indians diabetes data set is used as training and testing sets which consist of 768 data whereby 268 of them are diagnosed with diabetes. The result of this study will provide solutions to the medical staff in determining whether someone is the diabetes sufferer or not which is much easier rather than currently doing a blood test. The results show that the proposed method performs better than the other models such as multi layer perceptron (MLP), RBF and ANFIS structure.

Hierarchical Takagi-Sugeno type fuzzy system for diabetes mellitus forecasting
Arash Sharifi, Asiyeh Vosolipour, Mahdi Aliyari Sh, Mohammad Teshnehlab
Machine Learning and Cybernetics, 2008 International Conference on
2008Conference
Abstract:

This study has developed classical and hybrid controllers for control of magnetic levitation system. Sliding mode and PID controllers are proposed as a classical controllers and neural network based controller is used for controlling a magnetic levitation system. Adaptive neural networks controller needs plant`s Jacobain, but here this problem solved by sliding surface and generalized learning rule in case to eliminate Jacobain problem. The simulation results show that these methods are feasible and more effective for magnetic levitation system control.

Hybrid control of magnetic levitation system based-on new intelligent sliding mode control
M Aliasghary, M Teshnehlab, A Jalilvand, M Aliyari Shoorehdeli, MA Nekoui
Journal of Applied sciences
2008Journal
Abstract:

In this study integer genetic algorithm is applied for path planning of mobile robot in the grid form environment. The novel representation is proposed for definition of chromosome which reduced the computational complexity of genetic algorithm which was used before for path planning. Comparison with other encoding of chromosome is done to show the capability of proposed algorithm. Another genetic algorithm is used to repair some paths which collide with obstacles. Mamadani fuzzy rule is used to describe difficulty of passing from cells which are sandy or have slope.

Integer GA for mobile robot path planning with using another GA as repairing function
M Mansouri, M Aliyari Shoorehdeli, M Teshnehlab
Automation and Logistics, 2008. ICAL 2008. IEEE International Conference on
2008Conference
Abstract:

In this study, designing of multi-objective (MO) proportional, integral and derivative (PID) controller for load frequency control (LFC) based on adaptive weighted particle swarm optimization (AWPSO) has been proposed. Unlike single objective optimizations methods, MO optimization can find different solutions in a single run and we can select appropriate and desirable solution based on valuation to the objects. In this study for PID controller design, overshoot/undershoot and settling time are used as objective functions for MO optimization in LFC problem. So that various solutions with different overshoot/undershoot and settling time obtained. From these different PID parameters, one can select a single solution based on valuation to objects and as well as system constraints, reliability etc. The proposed method is used for designing of PID parameters for two area interconnected power system. From the simulation results, efficiency of proposed controller design can be seen.

Load frequency control in interconnected power system using multi-objective PID controller
K Sabahi, A Sharifi, M Aliyari, M Teshnehlab, M Aliasghary
Journal of Applied Sciences
2008Journal
Abstract:

In this paper designing of multi-objective PID controller for load frequency control (LFC) based on adaptive weighted particle swarm optimization (AWPSO) has been proposed. Conventional methods such as Ziegler-Nichols and Cohen-Coon are based on trial-and- error and their best performances are achieved for first-order process. Single-objective population based methods such as genetic algorithm (GA) and particle swarm optimization (PSO) have only one solution in a single run. Unlike single objective methods, multi- objective optimization can find different solutions in a single run. In the proposed method, overshoot/undershoot and settling time are used as objective functions for multi-objective optimization. The proposed method is used for designing of PID parameters for two area interconnected power system.

Load frequency control in interconnected power system using multi-objective PID controller
A Sharifi, K Sabahi, M Aliyari Shoorehdeli, MA Nekoui, M Teshnehlab
Soft Computing in Industrial Applications, 2008. SMCia'08. IEEE Conference on
2008Conference
Abstract:

Neural network Based controller is used for controlling a magnetic levitation system. Feedback error learning (FEL) can be regarded as a hybrid control to guarantee stability of control approach. This paper presents simulation of a magnetic levitation system controlled by a FEL neural network and PID controllers. The simulation results demonstrate that this method is more feasible and effective for magnetic levitation system control.

Magnetic levitation control based-on neural network and feedback error learning approach
M Aliasghary, M Aliyari Shoorehdeli, A Jalilvand, M Teshnehlab
Power and Energy Conference, 2008. PECon 2008. IEEE 2nd International
2008Conference
Abstract:

This paper introduces a new hybrid approach for training the adaptive network based fuzzy inference system (ANFIS) and a new type of particle swarm optimizers (PSO). The previous works emphasized on gradient base method or least square (LS) based method. This study applied one of the swarm intelligent branches, PSO. The hybrid method composes Fuzzy PSO with recursive least square (RLS) for training. We use PSO with some changes for training procedure parameters in antecedent part. These changes are inspired from Fuzzy Systems method and using Fuzzy rules for tuning PSO parameters during training algorithms. The simulation results show that in comparison with current gradient based training, and authors previous hybrid method the proposed training have a good adaptation to complex plants and train less parameter than gradient base methods.

Novel hybrid learning algorithms for tuning ANFIS parameters as an identifier using fuzzy PSO
Mohammad Teshnehlab, M Aliyari Shoorehdeli, Ali Khaki Sedigh
Networking, Sensing and Control, 2008. ICNSC 2008. IEEE International Conference on
2008Conference
Abstract:

In this study integer genetic algorithm is applied for path planning of mobile robot in the grid form environment. The novel representation is proposed for definition of chromosome which reduced the computational complexity of genetic algorithm that was used before for path planning. Comparison with other encoding of chromosome is done to show the capability of proposed algorithm. Mamadani fuzzy rule is used to describe difficulty of passing from cells which are sandy or have slope.

Path planning of mobile robot using integer ga with considering terrain conditions
Mohammad Mansouri, Mehdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Systems, Man and Cybernetics, 2008. SMC 2008. IEEE International Conference on
2008Conference
Abstract:

This paper, describes a hybrid control method to control a flexible joint. Dynamic equation of the system has been derived. The designed controllers consist of two parts: classical controller, which is a Linear Quadratic Regulation (LQR), and a hybrid controller, utilizing sliding mode control using Gaussian Radial Basis Function Neural Networks (RBFNN). The RBFNN is trained during the control process and it is not necessary to be trained off-line.

Sliding Mode Control of Flexible Joint Using Gaussian Radial Basis Function Neural Networks
F Farivar, M Aliyari Shoorehdeli, MA Nekoui, M Teshnehlab
Computer and Electrical Engineering, 2008. ICCEE 2008. International Conference on
2008Conference
Abstract:

This paper has developed a sliding mode controller (SMC) based on a radial basis function model for control of Magnetic levitation system. Adaptive neural networks controllers need plant's Jacobain, but here this problem solved by sliding surface and generalized learning rule in case to eliminate Jacobain problem. The simulation results show that this method is feasible and more effective for Magnetic levitation system control.

Sliding mode control of magnetic levitation system using radial basis function neural networks
Mortaza Aliasghary, Abolfazl Jalilvand, Mohammad Teshnehlab, M Aliyari Shoorehdeli
Robotics, Automation and Mechatronics, 2008 IEEE Conference on
2008Conference
Abstract:

This study suggests new learning laws for Adaptive Network based Fuzzy Inference System that is structured on the basis of TSK type III as a system identifier. Stable learning algorithms for consequence parts of TSK type III rules are proposed on the basis of the Lyapunov stability theory and some constraints are obtained. Simulation results are given to validate the results. It is shown that instability will not occur for learning rates in the presence of constraints. The learning rate can be calculated online from the input–output data, and an adaptive learning for the Adaptive Network based Fuzzy Inference System structure can be provided.

Stable Learning Algorithm Approaches for ANFIS As an Identifier
Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab, Ali Khaki Sedigh
World Congress
2008Conference
Abstract:

Particle swarm optimization (PSO) as a novel computational intelligence technique, has succeeded in many continuous problems. But in discrete or binary version there are still some difficulties. In this paper a novel binary PSO is proposed. This algorithm proposes a new definition for the velocity vector of binary PSO. It will be shown that this algorithm is a better interpretation of continuous PSO into discrete PSO than the older versions. Also a number of benchmark optimization problems are solved using this concept and quite satisfactory results are obtained.

A novel binary particle swarm optimization
Mojtaba Ahmadieh Khanesar, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
Control & Automation, 2007. MED'07. Mediterranean Conference on
2007Conference
Abstract:

In this paper a decoupled sliding-mode with fuzzy neural network controller for a nonlinear system is presented. To divided into two subsystems to achieve asymptotic stability by decoupled method for a class of three order nonlinear system. The fuzzy neural network (FNN) is the main regulator controller, which is used to approximate an ideal computational controller. The compensation controller is designed to compensate for the difference between the ideal computational controller and the FNN controller. A tuning methodology is derived to update weight parts of the FNN. Using Lyapunov law, we derive the decoupled sliding-mode control law and the related parameters adaptive law of FNN. The method can control one-input and multi-output nonlinear systems efficiently. Using this approach, the response of system will converge faster than that of previous reports.

Decoupled sliding-mode with fuzzy neural network controller for EHSS velocity control
Seyed Alireza Mohseni, Mahdi Aliyari Shooredeli
Intelligent and Advanced Systems, 2007. ICIAS 2007. International Conference on
2007Conference
Abstract:

In this study, a hybrid learning algorithm for training the recurrent fuzzy neural network (RFNN) is introduced. This learning algorithm aims to solve main problems of the gradient descent (GD) based methods for the optimization of the RFNNs, which are instability, local minima and the problem of generalization of trained network to the test data. PSO as a global optimizer is used to optimize the parameters of the membership functions and the GD algorithm is used to optimize the consequent part's parameters of RFNN. As PSO is a derivative free optimization technique, a simpler method for the train of RFNN is achieved. Also the results are compared to GD algorithm.

Hybrid training of recurrent fuzzy neural network model
Mojtaba Ahmadieh Khanesar, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab
Mechatronics and Automation, 2007. ICMA 2007. International Conference on
2007Conference
Abstract:

This paper present power system load frequency control by modified dynamic neural networks controller. The controller has dynamic neurons in hidden layer and conventional neurons in other layers. For considering the sensitivity of power system model, the neural network emulator used to identify the model simultaneously with control process. To have validation of proposed structure of neural network controller the results of simulation demonstrated that the proposed controller offers better performance than conventional neural network controller.

Load frequency control in interconnected power system using modified dynamic neural networks
K Sabahi, MA Nekoui, M Teshnehlab, M Aliyari, M Mansouri
Control & Automation, 2007. MED'07. Mediterranean Conference on
2007Conference
Abstract:

Use of multi-objective particle swarm optimization for designing of planar multilayered electromagnetic absorbers and finding optimal Pareto front is described. The achieved Pareto presents optimal possible trade offs between thickness and reflection coefficient of absorbers. Particle swarm optimization method in comparison with most of optimization algorithms such as genetic algorithms is simple and fast. But the basic form of multi-objective particle swarm optimization may not obtain the best Pareto. We applied some modifications to make it more efficient in finding optimal Pareto front. Comparison with reported results in previous articles confirms the ability of this algorithm in finding better solutions.

Modified multi-objective particle swarm optimization for electromagnetic absorber design
Somayyeh Chamaani, Seyed Abdullah Mirtaheri, Mohammad Teshnehlab, Mahdi Aliyari Shooredeli
Applied Electromagnetics, 2007. APACE 2007. Asia-Pacific Conference on
2007Conference
Abstract:

This paper introduces a new hybrid approach for training the adaptive network based fuzzy inference system (ANFIS). This approach based on multi objective optimization mechanism for training parameters in antecedent part. It considers two cost functions as the objectives which are the maximum difference measurements between the real nonlinear system and the nonlinear model, and training mean square error (MSE). The NSGA-II is the multi objective optimization algorithm which employed for this purpose. So we use gradient decent (GD) method for training all parameters in conclusion part. Finally we show simulation results of applied this method to some nonlinear identification system.

Multi objective optimization of ANFIS structure
V Seydi Ghomsheh, M Aliyari Shoorehdeli, A Sharifi, M Teshnehlab
Intelligent and Advanced Systems, 2007. ICIAS 2007. International Conference on
2007Conference
Abstract:

This paper introduces a new hybrid approach for training the adaptive network based fuzzy inference system (ANFIS). The previous works emphasized on gradient base method or least square (LS) based method. In this study we apply one of the swarm intelligent branches, named particle swarm optimization (PSO). The hybrid method composes PSO with recursive least square (RLS) for training. We use PSO with some changes for training procedure parameters in antecedent part. These changes are inspired from Genetic Algorithm (GA) method and using Adaptive Weighted for PSO. The simulation results show that in comparison with current gradient based training, the novel training can have a comparable adaptation to complex plants and train less parameter than gradient base methods. Also, the results show this new hybrid approach has less complexity than other gradient based methods.

Novel hybrid learning algorithms for tuning ANFIS parameters using adaptive weighted PSO
M Aliyari Shoorehdeli, Mohammad Teshnehlab, AK Sedigh
Fuzzy systems conference, 2007. FUZZ-IEEE 2007. IEEE international
2007Conference
Abstract:

This paper presents sliding mode control of rotary inverted pendulum. Rotary inverted pendulum is a nonlinear, unstable and non-minimum-phase system. Designing sliding mode controller for such system is difficult in general. Here, first the desired performance is introduced and based on this performance two sliding surfaces are designed, then system is controlled by proper definition of a lyapunov function. The lyapunov function designed puts more emphasis on the control of the inverted pendulum rather than the control of the motor.

Sliding mode control of rotary inverted pendulm
Mojtaba Ahmadieh Khanesar, Mohammad Teshnehlab, Mahdi Aliyari Shoorehdeli
Control & Automation, 2007. MED'07. Mediterranean Conference on
2007Conference
Abstract:

This paper introduces a new approach for training the adaptive network based fuzzy inference system (ANFIS). The previous works emphasized on gradient base method or least square (LS) based method. In this study we apply one of the swarm intelligent branches, named particle swarm optimization (PSO) with some modification in it to the training of all parameters of ANFIS structure. These modifications are inspired by natural evolutions. Finally the method is applied to the identification of nonlinear dynamical system and is compared with basic PSO and showed quite satisfactory results.

Training ANFIS structure with modified PSO algorithm
V Seydi Ghomsheh, M Aliyari Shoorehdeli, M Teshnehlab
Control & Automation, 2007. MED'07. Mediterranean Conference on
2007Conference
Abstract:

This study addresses new hybrid approaches for velocity control of an electro hydraulic servosystem (EHSS) in presence of flow nonlinearities and internal friction. In our new approaches, we combined classical method based-on sliding mode control and fuzzy RBF networks. The control by using adaptive networks need plant's Jacobean, but here this problem solved by sliding surface. It is demonstrated that this new technique have good ability control performance. It is shown that this technique can be successfully used to stabilize any chosen operating point of the system. All derived results are validated by computer simulation of a nonlinear mathematical model of the system. The controllers which introduced have big range for control the system.

Velocity control of an electro hydraulic servosystem
M Aliyari Shoorehdeli, Mohammad Teshnehlab, H Aliyari Shoorehdeli
Systems, Man and Cybernetics, 2007. ISIC. IEEE International Conference on
2007Conference
Abstract:

This paper introduces a new hybrid approach for training the adaptive network based fuzzy inference system (ANFIS). The previous works emphasized on gradient base method or least square (LS) based method. In this study we apply one of the swarm intelligent branches, named particle swarm optimization (PSO). The hybrid method composes PSO with gradient decent (GD) for training. We use PSO with some changes for training procedure parameters in antecedent part. These changes are inspired from genetic algorithm (GA) method. The simulation results show that in comparison with current GD training, the novel training can have a better adaptation to complex plants. Also, the results show this new hybrid approach optimizes ANFIS parameters faster and better parameters than gradient base method

A novel training algorithm in ANFIS structure
M Aliyari Shoorehdeli, M Teshnehlab, AK Sedigh
Proceedings of the American Control Conferences
2006Conference
Abstract:

This paper has introduced a new method for feature subset selection to which less attention has been given. Most of the past works have emphasized feature extraction and classification using classical methods for these works. The main goal in feature extraction is presented data in lower dimension. One of the popular methods in feature extraction is principle component analysis (PCA). This method and similar methods rely mostly on powerful classification algorithms to deal with redundant and irrelevant features. In this paper we introduced particle swarm optimization (PSO) as a simple, general, and powerful framework for selecting good subsets of features, leading to improved detection rates. We used PCA for feature extraction and support vector machines (SVMs) for classification. The goal is to search the PCA space using PSO to select a subset of eigenvectors encoding important information about the target concept of interest. Another object in this paper is to increase speed of convergence by using PSO to find the best feature. We have tested the framework in mind on challenging application like face detection. Our results illustrate the significant improvement in this case

Feature Subset Selection for face detection using genetic algorithms and particle swarm optimization
M Aliyari Shoorehdeli, Mohammad Teshnehlab, H Abrishami Moghaddam
Networking, Sensing and Control, 2006. ICNSC'06. Proceedings of the 2006 IEEE International Conference on
2006Conference
Abstract:

In this paper a novel hybrid strategy is employed in order to improve the controller performance. The main idea is combination of classical and intelligent controllers. Feedback error learning (FEL) as a two degrees of freedom (2DOF) control scheme, has been introduced based on this idea. This paper takes a step ahead of traditional FEL schemes which combine a PID controller with an intelligent inverse based controller. We introduce a robust FEL scheme and the robust controller replaces the conventional PID controller. The Robust controller is designed based on the Hinfin approach and the intelligent controller has ANFIS structure. This novel algorithm is implemented in a Flow plant to track the desired value of flow and reject unwanted disturbances in the practical system. The results are brought to prove the practical power of the novel method and are compared with other control schemes.

Flow Control Using a Combination of Robust and NeuroFuzzy Controllers in Feedback Error Learning Framework
R Adlgostar, Y Kouhi, M Teshnehlab, M Aliyari
Industrial Technology, 2006. ICIT 2006. IEEE International Conference on
2006Conference